Journal of Photochemistry and Photobiology C: Photochemistry Reviews 13 (2012) 28–54
Contents lists available at SciVerse ScienceDirect
Journal of Photochemistry and Photobiology C:
Photochemistry Reviews
journal homepage: www.elsevier.com/locate/jphotochemrev
Invited review
Studies on the interaction of pulsed lasers with plasmonic gold nanoparticles
toward light manipulation, heat management, and nanofabrication
Shuichi Hashimoto a,∗ , Daniel Werner a , Takayuki Uwada b
a
b
Department of Ecosystem Engineering, The University of Tokushima, Tokushima 770-8506, Japan
Department of Applied Chemistry and Institute of Molecular Science, National Chiao Tung University, Hsinchu 30010, Taiwan
a r t i c l e
i n f o
Article history:
Received 5 December 2011
Received in revised form 18 January 2012
Accepted 23 January 2012
Available online 2 February 2012
Keywords:
Gold nanoparticles
Laser-induced size reduction
Surface plasmon resonance
Photothermal response
Plasmonic nanobubble
a b s t r a c t
This review describes the fundamental aspects of laser–gold nanoparticle (Au NP) interaction that leads
to nanoscale energy deposition to the surroundings through light amplification and heat generation.
Besides the importance of the primary process in physics and chemistry, application of the light–NP
interaction has attracted significant interest from various areas ranging from analytical chemistry to
material chemistry and biomedicine. Here we consider both mechanistic and application aspects. Our
attention is focused on pulsed-laser-induced fast processes that revealed the heating–cooling dynamics
of electrons, lattice (particle), and particle’s environment. On the application side, we focus on material
fabrication and processing that beat diffraction-limited resolution. Together, we will shed a light on
the essence of research activities carried out in the past 10 years. In addition to an abundance of latest
information obtained from currently available literature, this review includes figures obtained by our
own calculations to provide readers with a better understanding of the basics of the optical properties
and energy and heat-transfer processes of Au NPs, which are not familiar to photochemists.
© 2012 Elsevier B.V. All rights reserved.
Contents
1.
2.
3.
4.
5.
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Brief background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.1.
LSPR: localized surface plasmon resonance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.2.
Electron dynamics, photothermal response and related phenomena . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.3.
Plasmonic field enhancement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Photo-induced dynamics of metallic nanoparticles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.1.
Shape transformation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.2.
Fragmentation and size reduction of gold nanoparticles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.2.1.
Possible mechanism of size reduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.2.2.
Temperature calculation to elucidate the mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.2.3.
Interpretation of experimental size reduction based on temperature simulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.2.4.
Time scale of fragmentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.2.5.
Single-particle approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.3.
Heat transfer from gold nanoparticles to the surrounding medium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.4.
Laser-induced agglomeration and fusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Application of laser-induced response of metallic nanoparticles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.1.
Surface modification and treatment of substrate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.2.
Laser-induced deposition and treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.3.
Miscellaneous applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Summary and future outlook . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
∗ Corresponding author. Tel.: +81 88 656 7389; fax: +81 88 656 7598.
E-mail address: [email protected] (S. Hashimoto).
1389-5567/$20.00 © 2012 Elsevier B.V. All rights reserved.
doi:10.1016/j.jphotochemrev.2012.01.001
29
30
30
31
33
33
33
35
35
36
38
40
42
42
45
46
46
49
49
51
52
52
S. Hashimoto et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 13 (2012) 28–54
Shuichi Hashimoto has been a Professor of Ecosystem
Engineering, at the Graduate School of Technology and
Science, The University of Tokushima, Japan since 2005.
He received a PhD in Physical Chemistry from Tokyo
Metropolitan University in 1982 and carried out postdoctoral research at the University of Notre Dame under
J.K. Thomas from 1982 to 1985. He worked with H.
Masuhara as a visiting scholar at Osaka University, investigating the transient-absorption spectroscopy of zeolites
in 1993. His current research interests are laser-induced
size and shape transformations of plasmonic nanoparticles, microscopy–spectroscopy study of nanoparticles and
nanostructures, femtosecond-laser processing of transparent solids, and optical microscopy study of zeolite photochemistry.
Daniel Werner, native of Dresden, Germany, is currently
a JSPS postdoctoral fellow at The University of Tokushima.
He received PhD in Physical Engineering from The University of Tokushima working with Prof. S. Hashimoto. His
current research interest focuses on the nonlinear optical properties of multilayered core–shell nanoparticles,
electron dynamics and thermodynamics of laser-heated
nanoparticles in liquids, fragmentation mechanisms of
noble metal nanoparticles and nanoparticle generation in
liquids through a pulsed-laser-ablation technique.
Takayuki Uwada received a PhD in Applied Physics from
Osaka University under the guidance of Prof. T. Asahi and
Prof. H. Masuhara in 2007. After a year of postdoctoral
research at the Hamano Life Science Research Foundation
in Kobe, he has been appointed as an assistant research
fellow in the Department of Applied Chemistry and Institute of Molecular Science, National Chiao Tung University,
Taiwan under the guidance of Prof. Masuhara. His current
research interests include static and dynamic lightscattering-microspectroscopy investigations of nanomaterials, laser-induced molecular/nanoparticle assembly
and crystallization, and metallic nanoparticle plasmon
application to biophysics.
1. Introduction
Past decades have witnessed an ever increasing interest in the
optical properties of nanoparticles and nanostructures [1–7], which
has had a tremendous impact on broad areas of research such as
optics, electronics, biomedicine, analytical chemistry, and photochemistry. Scheme 1 gives an overview of the areas of current
promising applications of plasmonic nanoparticles (NPs).
In nano-optics and nanoelectronics, miniaturization with high
throughput is highly desirable; consequently, an interest in surface plasmons in optical circuits below the diffraction limit has
grown rapidly [8,9]. In biomedical applications, because of their
strongly resonant light-absorbing and light-scattering properties,
Scheme 1. Overview: the areas of current promising applications of plasmonic
nanoparticles.
29
plasmonic NPs can be useful contrast agents in the diagnostic imaging of tumors [10,11]. In addition, when illuminated, plasmonic
NPs can serve as nanoscale heat sources, photothermally inducing
cell death and tumor remission [12,13]. In chemistry, in particular for analytical applications, surface-enhanced Raman scattering
spectroscopy (SERS) utilizing enhanced local electromagnetic fields
near the plasmonic NPs and nanostructures is a promising tool
for ultratrace analysis, ultimately enabling detection at the singlemolecule level [14,15]. Biosensing that exploits the detection of
a refractive-index change around the plasmonic NPs is another
area of development [16–18]. Recent developments have greatly
improved the sensitivity of plasmon sensors based on the single
NPs and NP arrays. In the plasmonic NP field, photochemistry is
currently prominent among the areas receiving increasing attention.
Photovoltaics, which is a method of converting sunlight into
electricity, is a major field in photochemistry [19–21]. Approaches
based on plasmonic structures can be used to improve absorption
in photovoltaic devices, permitting a considerable reduction in the
thickness of absorber layers and yielding new options for solar
cell design [22,23]. An effect of plasmon-assisted optical antennas
[24,25], where light is concentrated into a subwavelength volume to convert the incident light into amplified localized fields,
has the potential to promote low-yield photochemical reactions
with greater efficiency. It has been shown that two-photon-induced
photochromic and photopolymerization reactions can be realized
under the illumination of continuous wave (CW) lasers and halogen
lamps [26–29]. Previously, these two-photon reactions have only
been conducted by irradiation with high-intensity pulsed lasers.
Fundamental aspects and applications of metal-enhanced fluorescence [30,31] and plasmon-assisted photocatalysts [32,33] are also
active areas of research in photochemistry.
This review deals with phenomena initiated by the interaction
of lasers with plasmonic NPs that give rise to rich physics and chemistry, as summarized in Scheme 2.
Briefly, the absorption of visible light by a plasmonic NP results
in the instantaneous heating of the particle [34,35]. This heating
has a profound effect both on the particle itself and the surrounding medium. The former can be described by electron dynamics
[4–7], coherent acoustic lattice vibrations [6,36,37], melting and
evaporation from the particle surface [38,39], and explosive fragmentation [40,41]. The latter is described by cavitation [42–45] and
stress-wave generation [46,47], resulting from heat transfer from
the particle to the surroundings. These phenomena are intriguing
from a fundamental scientific viewpoint and cover a wide range of
applications as will be described below (Section 4). For instance,
local heating triggered by the light absorption of a plasmonic NP
enables nanofabrication on particle support. The amount of heat
generated by the plasmonic NP can be controlled by the particle
size, shape, and aggregation state of the NP as well as the illuminating laser intensity, wavelength, and pulse duration. Thus interplay
of plasmonic NPs with laser light allows spatial and temporal heat
management besides the optimization of local electromagnetic
field.
This review is intended to provide photochemists with information that is useful when initiating research using plasmonic NPs, in
particular with lasers combined with traditional organic and inorganic photochemistry. The review focuses on metal NPs rather than
semiconductor quantum dots [48,49], which are more familiar to
photochemists. The photochemistry of noble metal nanoparticles is
different from that of traditional molecules. Generally, the fluorescence emission of Au NPs is very weak unless the particle diameters
are less than 5 nm [50]. On the other hand, Au NPs with diameters greater than 20 nm scatter light in the dark field and can be
used for imaging just like fluorescent dyes or quantum dots without suffering from photobleaching/blinking [51,52]. Their particle
30
S. Hashimoto et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 13 (2012) 28–54
Scheme 2. Interaction of pulsed lasers with plasmonic nanoparticles and related phenomena.
size-, shape-, spacing- and medium-dependent optical properties
have distinct advantages that cannot be replaced by any other
materials. Note, however, that the accepted method of describing
the optical properties of these metal NPs is quite different from that
by which the photophysics of other molecules is delineated. This is
because, for metals, frequency (wavelength)-dependent responses
of dielectric functions are important. In this review, we examine
the emerging field of laser–plasmonic NP interactions, leading to
light manipulation, heat management, and nanofabrication. Particular emphasis is placed on Au NPs interacting with pulsed lasers.
Although Ag NPs have superior characteristics in terms of plasmonic enhancement [53], Au NPs have been selected because
they have higher chemical stability and are used more frequently
[54,55].
2. Brief background
2.1. LSPR: localized surface plasmon resonance [1–6]
Plasmonic NPs absorb and scatter light from UV to near IR. The
classical interaction of NPs with light and the resultant optical
properties are described by the dielectric function ε; here ε is usually represented as a function of the angular frequency ω (=2c/)
of light. The dielectric function of gold is defined as the sum of the
interband term, εIB (ω), which considers the response of 5d electrons to the 6sp conduction band, and the Drude term, εD (ω), which
considers free conduction electrons (Drude model); thus,
ε(ω) = εIB (ω) + εD (ω).
(1)
The localized surface plasmon resonance (LSPR), representing the optical properties characteristic of gold NPs, originates
from the latter term, i.e. the photoexcitation of free electrons.
The collective oscillation of photoexcited free electrons on the
surface of Au NPs occurs via the interaction of Au nanospheres
with the electromagnetic field of light, exhibiting a resonance
frequency at 510–550 nm depending on the particle radius
(extrinsic size effect). By contrast, the interband transition is
insensitive to the particle size, yielding a threshold at 518 nm
(2.4 eV). Each term of the dielectric function is divided into a
real part (ε1 (ω) = n2 − 2 ) and an imaginary part (ε2 (ω) = 2n),
where n represents the refractive index and represents
S. Hashimoto et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 13 (2012) 28–54
the absorption coefficient of the metal (complex refractive index:
n = n + i).
Plasmonic NPs with diameters lesser than the electron-meanfree path (∼40 nm for Au) allow the confinement of free electrons.
According to the simplest model, the electrons are damped by a
inelastic scattering process at the inner NP surface, resulting in
an increased relaxation frequency or a higher damping constant
(s−1 ): the relaxation time of the free-electron gas is reduced.
Accordingly, the dielectric function (ε(ω)) of a bulk metal has to
be modified (intrinsic size effect), giving rise to a size-dependent
dielectric function ε(ω, R),
ε (ω, R) = εbulk (ω) − εbulk
(ω) + εD (ω, R)
D
(2)
Here εbulk (ω) describes the dielectric function of bulk metal
calculated from n and , where values have been experimentally
obtained [56]. The second term in equation (2) denotes the Drude’s
dielectric function of a free-electron gas defined by equation (3a),
and the third term denotes the dielectric function of a NP with a
radius R defined by equation (3b). Inspection of equation (2) reveals
that the dielectric function of the free-electron gas εbulk (ω) determined experimentally is modified by including a size-dependent
term that considers the intrinsic size effect of NPs, technically by
subtracting that of the bulk metal εbulk
(ω) and adding a modified
D
dielectric function εD (ω,R):
εbulk
(ω) = 1 −
D
εD (ω, R) = 1 −
ωp2
(3a)
ω2 + iω
ωp2
(3b)
ω2 + i (R) ω
where ωp (s−1 ) represents the plasma frequency of the free1/2
electron gas defined by (ne e2 /ε0 m∗e )
= 1.38 × 1016 s−1 for Au,
where ne is the density of the free electrons (5.9 × 1028 m−3 for
Au), e is the electron charge (A s or C), ε0 is the vacuum permittivity
(A2 s4 kg−1 m−3 ) and m∗e is the effective mass of an electron (kg). The
damping constant is described by = vF /l ( = 1.07 × 1014 s−1
for Au) for the bulk metal and (R) = vF (1/l + A·1/R) for metal
nanospheres with diameters smaller than the electron-mean-free
path, where vF (1.4 × 106 m s−1 for Au) is the Fermi velocity, l
(3.8 × 10−8 m for Au) is the electron-mean-free path, R (m) is the
NP radius, and A is a proportionality constant (A = 1 for isotropic
electron scattering).
The absorption and scattering properties of spherical metal
NPs are well described by applying the Mie theory [57]. The Mie
theory provides exact analytical solutions to Maxwell’s equations
for spheres with various diameters: the absorption and scattering
probabilities are represented by the absorption cross section Cabs
(m2 ), the scattering cross section Csca (m2 ), and the extinction cross
section Cext , which is the sum of both (Cext = Csca + Cabs ). The cross
section, C, of each optical event is defined by
I = I0 exp(−N · C · x)
(4)
where I0 represents the intensity of incident light, I represents the
intensity of transmitted light, N (m−3 ) is the number density of
particles, and x (m) is the optical pass length. Equation (4) basically expresses the Beer-Lambert law. Within the framework of the
Mie theory, the absorption and scattering of light originate from
free-electron oscillations driven by the oscillating electromagnetic
fields of the incident light. As the diameter of the metal sphere
increases, oscillations from higher orders, such as quadrupole or
octapole oscillations, are likely to occur as well as the dipole mode.
These oscillations are currently defined as plasmon modes. In the
Mie calculation, the input parameters are the particle’s dielectric
function ε1 (ω) + iε2 (ω), radius R, and the permittivity of surrounding medium εm . The solution procedure is complicated, and thus
not given here [58]. In the quasi-static regime where the radius R is
31
significantly smaller than the incident wavelength , phase retardation and the mixing of higher-order multipole oscillations are
negligible. In this case, a more simplified expression based on the
dipole approximation can be used, as given by
3/2
Cext =
εm ε2 (ω)
12ωR3
c0
[ε1 (ω) + 2εm ]2 + ε2 (ω)2
Csca =
1285 R6 ε2m
34
(ε1 (ω) − εm )2 + ε2 (ω)2
(5a)
(ε1 (ω) + 2εm )2 + ε2 (ω)2
(5b)
Here, c0 is the velocity of light in vacuum. The resonance condition in which the peak of the LSPR band occurs is approximately
fulfilled when ε1 (ω) = −2εm , if ε2 (ω) is small or slightly dependent
on ω. Note, however, that equations (5a) and (5b) cannot reproduce the size-dependent peak shift of the LSPR band because of the
limitation of the dipole approximation.
Fig. 1 depicts the calculated spectra of extinction, absorption
and scattering for a single spherical Au NP of various diameters.
The distinct LSPR band appears at 520–580 nm. A rising shoulder with decreasing wavelengths at less than 500 nm is assigned
to the interband transition. The size-dependent optical properties are very obvious. Practically, the scattering of a 20-nm Au
NP is negligibly weak. A somewhat broadened spectral envelope
was obtained for an extinction (absorption) spectrum owing to the
intrinsic size effect. Furthermore, the LSPR band exhibits a much
broader envelope for particles with much smaller than 10 nm and
even disappears completely for NPs with sizes less than or equal
to 2 nm. The contribution of scattering is greater with increasing
diameter; for 100-nm Au particles, the relative amount of scattering
cannot be ignored. Higher-order modes are more prominent with
increasing particle size, causing the LSPR band to undergo a redshift and resulting in an increased bandwidth. This is because, for
large particles, the light cannot polarize the NPs homogeneously,
and retardation effects lead to the excitation of higher-order modes.
The near-field scattering cross section is approximately 2 orders
of magnitude greater than the scattering cross section for 20-nm Au
NPs because of higher-order electric fields located on the NP surface. The fields vanish rapidly with distance. In contrast, the CNF of
a 100-nm-diameter Au sphere is only 1 order of magnitude greater
than Csca .
The experimental spectra in colloidal solutions generally agree
well with the calculated spectra under the conditions of homogeneous particle size distribution and low concentrations. Besides,
single-particle absorption and scattering spectra that are directly
compared with the Mie theory can be measured [59,60]. Discrete dipole approximation method was performed to calculate
the extinction and scattering efficiencies of non-spherical particles,
rods, cubes, and triangular plates [61–63].
The LSPR band undergoes temperature-dependent spectral
changes. Fig. 2 shows the extinction spectral changes of a Au NP
in air on a glass substrate at various temperatures, calculated by
the Mie theory using the temperature-dependent dielectric functions measured for the gold film by Otter [64]. For comparison, an
experimental spectral change from literature is also shown [65].
With increasing temperature, the bleaching (intensity reduction) of the LSPR band occurs with the broadening of the spectral
envelope. This can be explained by an increase in the occurrence
of electron–phonon scattering at high temperatures. In this event,
the reduction in the ε2 (ω) values occurs and leads to an increased
damping constant .
2.2. Electron dynamics, photothermal response and related
phenomena
The ultrafast electron–phonon dynamics of Au and Ag
NPs have been well investigated by applying femtosecond
32
S. Hashimoto et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 13 (2012) 28–54
a
b
c
Fig. 1. Cross sections of extinction Cext , absorption Cabs , scattering Csca , and near-field scattering CNF , representing scattering on the NP surface, as a function of the wavelength
calculated by the Mie theory for diameters of 20 nm Au (a), 50 nm Au (b), and 100 nm Au (c) single spherical particles submerged in water (n = 1.33). Note that the ordinate
scales are different: the scales in (b) are 10 times greater than those of (a) and the scales in (c) are 10 times greater than those in (b).
Fig. 2. (a) Computational temperature-dependent extinction spectral changes of a 60-nm Au NP placed on a glass substrate in air (effective medium refractive index: 1.1).
(b) Temperature dependence of the plasmon band absorption (extinction) for the 22-nm Au NPs in water. The absorption spectra are measured at 291 K or 18 ◦ C (solid line)
and 345 K or 72 ◦ C (dashed line). (Fig. 2b was reproduced from Ref. [65]. Reproduced with permission from the American Chemical Society.)
transient-absorption spectroscopy [66–68]. The energy deposition
resulting from the laser interaction of plasmonic NPs occurs in the
following four stages [35,69]. First, the laser light is absorbed by the
electronic system, creating nascent nonthermal electrons. After a
few 100 fs, a thermal equilibrium is reached corresponding to the
Fermi distribution as a result of electron–electron scattering with
cold conduction-band electrons. Subsequently, the electron energy
is transferred to the NP lattice via electron–phonon coupling (<a
few ps). The lattice energy is later transferred to the surrounding medium through phonon–medium interaction (<several tens
to several hundreds of ps) through heating the medium at the
expense of the cooling NP. The final stage is heat diffusion within
the medium. Fig. 3 represents the electron relaxation dynamics of
a Au NP as a function of time on excitation by a femtosecond laser
[70].
The well-defined electronic temperature Te , resulting from
the internal thermalization of hot electron gas varies from hundred to several thousands of Kelvin depending on the laser
pump energy and pulse duration. The lattice temperature TL
increases as a result of the energy transfer from the hot
electron gas into the lattice; in the last step, the temperature of the surrounding medium Tm increases owing to the
energy transfer from phonons. These consecutive processes are
described well by applying the two-temperature model (TTM)
[4–7,34,69–72].
The heat transfer from the NPs to the medium leads to the formation of a photothermal bubble in liquids. It is believed that heat
dissipation from pulsed-laser-heated metal NPs causes an explosive evaporation of the surrounding superheated liquid, which
changes into an expanding bubble formation [42,43]. Scheme 3
illustrates the process of bubble generation caused by the pulsedlaser-induced heating of a Au NP and subsequent heat dissipation
to the surrounding liquid.
Fig. 3. Simulated temporal evolution of electron temperature Te (dashed red curve);
lattice temperature TL (solid black curve); and maximum water temperature Tm
at the NP–water interface, (dashed–dotted blue curve) for a 55-nm-diameter gold
sphere absorbing a 150-fs laser pulse (full-width at half maximum; FWHM of
the Gaussian time profile) at 400 nm and a laser energy density of 10 mJ cm−2
(Pmax = 6.3 × 1010 W cm−2 ). Horizontal lines represent temperatures at which important events occur: Tbp : boiling point of bulk gold (∼3100 K); Tmp : melting point of
bulk gold (1336 K); Tcp : the critical point of water (647 K); and Tcav : cavitation threshold (573 K, Ref. [73]), considering the spinodal effect of water. (Modified from Ref.
[70]. Reproduced with permission from the American Chemical Society.)
S. Hashimoto et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 13 (2012) 28–54
33
Scheme 3. Schematic illustration of explosive bubble generation in liquids through pulsed-laser-induced heating of a Au NP. The initial high-pressure shock wave caused
by the fast lattice expansion of the heated NP, which passes through the water with approximately the velocity of sound, is not considered here.
2.3. Plasmonic field enhancement [74,75]
The excitation of the large optical polarization associated with
LSPR results in large local electromagnetic field enhancement at
the NP surface, in addition to strongly enhanced absorption and
scattering by the NP at the LSPR frequency. An example for single
spherical Au particles, 50, 100, and 400 nm on excitation of the LSPR
band of Au is given in Fig. 4.
A greater enhancement is expected for particle diameters
40–60 nm and the optimum diameter results from both the absorption and scattering cross sections. This is clearly seen in Fig. 4, where
the field enhancement of the Au NP is greater for a 50-nm-diameter
particle than for 100-nm and 400-nm-diameter particles, because
of the greater contribution of scattering for bigger Au NPs. The electric field is strongly localized for small spheres. Contrastingly, the
dipole field of a 100-nm Au NP is distorted because of the contribution of multipole oscillations within the sphere. Further severe
distortion was observed for the 400-nm Au sphere, and higherorder plasmon modes are well visible. Importantly, the intrinsic
size effect (see Section 2.1) is included in the field-enhancement
calculation. This is quite important, if the Au NP radius is less
than or equal to the electron-mean-free path. For instance, by
neglecting the intrinsic size effect, the electric field-enhancement
factor increases to 6.04 for 50-nm Au NPs. The difference is more
dramatically observed for 20-nm-diameter Au NPs, where the fieldenhancement factors are 4.58 with the intrinsic size effect and
5.36 without it. Greater field enhancement was postulated on the
basis of experimental studies for non-spherical particles and nanostructures, including nanorods [76] and triangles [77] with sharp
edges, for which field calculations on the basis of finite-differencetime-domain (FDTD) simulations supporting the observations were
given [75–79].
The near-field scattering efficiency, defined by QNF = CNF /R2 ,
is a measure of the ability of a sphere to extract power from the
incident wave and redirect it as scattering power over the whole
surface. QNF is considerably greater than unity. This implies that
the influence of the sphere on the incident light extends beyond
its geometrical boundaries. In other words, a spherical Au NP can
act as a field amplifier. QNF is an important parameter for surfaceenhanced Raman scattering (SERS). Fig. 5a shows QNF as a function
of the diameter of Au NPs surrounded by water, calculated both at
532 nm and at the peak positions of the CNF .
When excited at 532 nm, the value of QNF increases with increasing NP diameter, reaching a plateau at diameters between 50 and
60 nm. This behavior can be explained by a greater increase in the
dependence of scattering cross section on the NP size because of the
reduced contribution of the intrinsic size effect. Further increase in
diameter leads to a reduced near-field scattering efficiency because
the field retardation and geometrical cross section start to dominate. By contrast, the curve given by the peak position of QNF is
shifted to larger NP sizes. This is reasonable because a red-shift
occurs for the resonance wavelength of the CNF . For instance, the
wavelength at the maximum peak position of CNF is 549 nm for a 50nm-diameter Au NP and 587 nm for a 100-nm Au NP. The near-field
scattering efficiency can be significantly larger than the scattering
efficiency Qsca , which is a measure of the far-field scattering. This
is because QNF involves fields that increase more rapidly than the
dependence of 1/a (a is the distance from the surface of a metallic
sphere) as one moves from the far field to the surface of the sphere.
Thus, the integrated near-field intensity includes additional field
components that appear only in the near-field region. Fig. 5b gives
the values of QNF as a function of a for a 50-nm Au NP dispersed
in water. The figure shows that QNF decays very quickly within a
distance of 25 nm, where QNF decreases to 7.8% of the maximum
value. After a distance of 50 nm is reached, QNF is only 2.1% of the
surface value. This value is nearly equal to Qsca . At this distance, the
far-field effect dominates.
3. Photo-induced dynamics of metallic nanoparticles
3.1. Shape transformation
Pulsed-laser irradiation of Au nanorods suspended in water
resulted in a shape change to spherical particles [38,39,80–82]. The
shape transition of the nanorods was studied by Chang et al. [80]
by exposure to 532- and 1064-nm nanosecond (6 ns) lasers. They
ascribed the observed rod-to-sphere conversions to a photoannealing process. Link and co-workers carried out a similar experiment
by applying both femtosecond (800 nm, 100 fs) and nanosecond
(355 nm, 7 ns) lasers [38,39,81]. They observed that both lasers
basically contribute to the photothermal reshaping of Au nanorods
in an aqueous solution. They also found that femtosecond lasers
are better suited for reshaping purpose because the energy threshold for complete melting of the nanorods is reduced by a factor of
100 when using the femtosecond laser. This is because femtosecond lasers have such a high intensity that the rate of absorption,
and thus heating the lattice, is extremely rapid. This point will be
described later with reference to temperature simulation.
Zijlstra et al. [82] observed a gradual shape change from a long
and thin rod to a shorter and wider rod, which eventually collapsed into a sphere, thus ambiguously confirming the assumption
of the rod-to-sphere transformation actually occurring (Fig. 6). In
this case, a simple calculation showed that the mp of gold (1337 K)
is attained for the complete transformation by the laser energy
deposited (260 fJ or 2.1 mJ cm−2 ) for nanorods with dimensions
92 × 30 nm under the condition of negligible heat dissipation.
It has been known that NPs exhibit a melting temperature that
depends on particle size because of the effect of the surface-tovolume ratio. A remarkable decrease in the melting temperature
was observed for particle diameters less than 5 nm [83–85]. For
34
S. Hashimoto et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 13 (2012) 28–54
Fig. 4. Electric field (|E|/|Ei |) distributions for diameters of 50 nm (a), 100 nm (b), and 400 nm (c) Au spheres submerged in water (n = 1.33) on excitation by a plane-polarized
light with a wavelength of 532 nm. The calculation was performed on the basis of the Mie theory. The k-vector of the incident field Ei is collinear in the z-direction and the
electric field oscillates in the x–z plane. The field in the y–z plane was not given because of insignificant enhancement. Note that the light intensity is the square of the field
strength, and thus, the enhancement of incident light intensity should be (|E|/|Ei |)2 .
particles lesser than 5 nm, the homogeneous melting temperature
is inversely proportional to the radius of the particle, whereas for
larger particles, melting occurs near the mp of bulk gold. Experimentally, the melting temperature of a 38-nm Au NP has been
found to be ∼1067 K (∼830 ◦ C), which is 200 K lesser than that of
bulk gold with a mp of 1337 K (1064 ◦ C) [86]. Distinct from the
entire particle melting at above the melting temperature, it was
demonstrated that shape changes induced by surface melting occur
at temperatures much lower than the mp of bulk gold. For instance,
a transformation in the shape of average 38-nm elliptical Au NPs
resulting from surface melting was observed at temperatures as
low as 673 K (400 ◦ C) when the particles deposited on a silicon substrate were heated in a muffle furnace and inspected using SEM
a
[86]. Thermal heating experiments affirmed that a temperature of
523 K (250 ◦ C) is sufficient to completely transform nanorods into
spheres within an hour [87]. In this case, reshaping was observed
even at 373 K (100 ◦ C). A phase transformation ascribable to the formation of a core–shell structure of a particle with a surface melting
layer was experimentally observed at 377 K (104 ◦ C) [88]. These
results suggest that surface melting occurs at a temperature much
lower than the mp of bulk gold.
Pulsed-X-ray scattering was used to examine the lattice expansion and melting dynamics in Au NPs (100-nm diameter) in water,
following excitation with femtosecond-laser pulses (excitation
wavelength: 400 nm) [89]. At the excitation power corresponding to temperatures below the mp of bulk gold, the lattice shows
b
a
Fig. 5. (a) Near-field scattering efficiency QNF as a function of the diameter of Au NP surrounded by water, calculated both at 532 nm and at the peak positions of the CNF ; (b)
QNF as a function of a (the distance from the surface of a metallic sphere) for a 50-nm Au NP dispersed in water.
S. Hashimoto et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 13 (2012) 28–54
Fig. 6. Sketch of the intermediate particle shapes observed at different pulseenergy densities. The dimensions of all scanning electron microscopy (SEM) images
are 200 × 150 nm. The laser irradiation was conducted by a single pulse of 100 fs
duration with a wavelength of 800 nm. (Reproduced from ref. 82. Reproduced by
permission of the PCCP Owners Society.)
a loss of long-range order owing to the premelting of the particle. When the power was increased to the melting temperature
of bulk gold, complete melting occurred within 100 ps after laser
excitation. Because of the limitation of the time resolution of
the system, the detection of earlier events was not possible.
Besides, one would have to differentiate between particle melting of the particle ensemble and that of the structure of each single
particle.
3.2. Fragmentation and size reduction of gold nanoparticles
3.2.1. Possible mechanism of size reduction
Exposure of Au and Ag NPs to a single-shot laser or a number of
shots of pulsed lasers allowed the size reduction or fragmentation
[38,39,70,90–100]. The laser-induced size reduction has attracted
significant attention from researchers in the related fields. This phenomenon includes the fundamental aspects of interplay between
NPs and lasers such as how photon energy flows within NPs, leading to the destructive event, and how fast the energy flow occurs. In
addition, the method can be utilized for controlling particle size and
the size distribution in the laser-ablation-based nanoparticle generation [101–105]. Three types of mechanisms have been known
to consider size reduction, as illustrated in Scheme 4 [70].
Koda and co-workers [90] proposed a heating–melting–
evaporation concept (Scheme 4 (1)) for the observation of the size
reduction of chemically prepared aqueous Au NPs with sizes ranging from 19 to 47 nm diameter. In their experiments, the LSPR
band was exposed to various intensities and numbers of a 532-nm
nanosecond-pulsed-laser beam. Assuming negligible heat transfer
to the surrounding water, they considered that the size reduction
starts to occur when the particle temperature exceeds the boiling
point of bulk gold. The calculation of the particle temperature T (K)
was performed by applying simple thermodynamics:
T [K] =
Q − Hmelt − Hvap
+ 293
Cp
(6)
Here Q (J g−1 ) represents the energy absorbed by a unit mass of
gold by the irradiation of a single pulsed-laser light, Cp is the specific
heat capacity of Au, 0.131 (J g−1 K−1 ) and Hmelt is the enthalpy of
melting, 6.28 × 101 (J g−1 ), and Hvap is the enthalpy of vaporization, 1.87 × 103 (J g−1 ). In support of the photothermal mechanism,
Inasawa et al. [91] conducted further detailed research through the
35
use of a 355-nm picosecond laser. They revealed that the size reduction of Au NPs takes place by a layer-by-layer mechanism based on
the observation of the bimodal distribution of particle sizes that are
slightly smaller and much smaller than the original ones. They concluded that the irradiation by picosecond-laser pulses is the most
efficient way to induce the size reduction of Au NPs. Later, Pyatenko
et al. [92] carried out a calculation to examine the melting and
size reduction of gold NPs and concluded that the photothermal
mechanism prevails at low laser intensities. He also concluded that
the excitation of the LSPR band of Au NP by a 532-nm nanosecond
laser is more efficient because of a greater absorption cross section
Cabs , compared with other wavelengths, such as 1064 and 355 nm,
but gave no experimental evidence. Werner et al. carried out the
measurement of the dependence of the size-reduction threshold
on the excitation wavelength [93]. They found that the interband
excitation is more efficient than the LSPR excitation.
In contrast to the photothermal evaporation concept, a Coulomb
explosion model is given in Scheme 4 (2). The Coulomb explosion model presumes the ejection of quite a large number of
electrons to generate multiple ionized NPs that undergo spontaneous fission because of the charge repulsion. Kamat et al.
adopted this model to interpret the 355-nm picosecond-laserinduced splitting of Ag NPs, because they detected hydrated
electrons in the transient-absorption measurement [94]. Subsequently, Mafune and co-workers observed hydrated electrons on
intense nanosecond-pulsed-laser excitation of Au NPs in aqueous
sodium dodecyl sulfate (SDS) [95,96]. Their results indicated higher
size-reduction efficiencies for 355-nm irradiation than for the
excitation of the LSPR band of the gold NPs. The observed high efficiency was explained by interband electronic excitation–relaxation
cycles that enable a faster absorption of a new photon than
the conduction-band electrons can absorb. Furthermore, they
detected positively charged Au ions by applying mass spectrometry and proposed that thermionic emission of electrons due to
the interband excitation is responsible for the highly efficient
splitting of Au NPs [97,98]. More recently, Giammanco et al.
developed a theoretical analysis based on the TTM to explain
the fragmentation of gold NPs by picosecond-laser irradiation
[99]. They excited rather small gold NPs with an average diameter of 3.5 nm stabilized with poly(amidoamine) PAMAM-G5 in
water. By fitting their experimental data with the theoretical
analysis, they concluded that the evaporation of the NPs does
not play a relevant role in comparison to the Coulomb explosion resulting from thermionic electron emission and photoelectric
effect. However, they did not consider the depletion of mp and
bp and the quantum confinement effect for such small particles.
The photothermal and Coulomb explosion mechanisms have
been investigated independently and no unified concepts have
been developed to distinguish the two mechanisms. For instance,
it remained unclear under what conditions of pulse width and
laser energy one mechanism surpasses the other. In the Coulomb
explosion mechanism, the electron dynamics that govern various
relaxation processes including electron emission, play a major role
in electronic excitation. In contrast, in the photothermal evaporation mechanism, evaporation from the particle surface plays a
major role. In this case, the thermal energy deposited into the
lattice system results in the heating of NPs to the evaporation
point depending on the evaporation pressure of the material. Until
recently, precise theoretical calculations to determine the thresholds for surface evaporation and Coulomb explosion have been
unavailable, especially for the nanosecond-pulsed-laser excitation
of NPs. Table 1 summarizes the previous studies of pulsed-laserinduced size reduction. Not many studies gave the experimental
threshold of size reduction. The threshold is so important to identify
the precise mechanism.
36
S. Hashimoto et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 13 (2012) 28–54
Scheme 4. Schematic illustration of size reduction and fragmentation mechanisms of plasmonic NPs in a colloidal solution.
In addition to the photothermal evaporation and Coulomb
explosion mechanisms, a near-field ablation mechanism [106]
has been postulated exclusively for extremely high-intensity
femtosecond-laser excitation (Scheme 4 (3)). The mechanism
assumes that the spallation of solid Au NPs takes place by the action
of an intense laser field that was caused by the plasmonic nearfield enhancement. The experimental proof for the mechanism is
still not convincing because of the following reason. The near-field
ablation of Au NPs in water is supposed to take place well before
the melting of the nanorods (NRs). However, El-Sayed’s group [81]
and Zijlstra et al. [82] have shown that Au NRs can easily melt to
form spherical NPs (see Section 3.1) without undergoing fragmentation. Au NRs have a much higher field-enhancement factor (see
Section 2.3) than spherical NPs. This may cause near-field ion emission leading to the Coulomb explosion at much lower laser densities
than those observed by Plech’s group [106]. Thus the concept of
near-field ablation contradicts the observations of melting without
fragmentation. Werner’s group performed field-intensity calculations on the surface of 60-nm Au NPs in water and concluded that
the field strength at the fragmentation threshold is not sufficiently
high for ion emission [70].
3.2.2. Temperature calculation to elucidate the mechanism
Werner and Hashimoto [100] conducted a precise calculation of time-dependent temperature profiles to elucidate the
size-reduction mechanism of aqueous colloidal Au NPs, which
is dependent on excitation laser-pulse duration (femto-, picoand nanosecond), excitation laser energy, and excitation laser
wavelength. They conducted a simulation based on both a TTM
[4–7,34,71,72,107–110] describing electron-lattice heating and the
liquid-droplet model of fragmentation [111,112] to distinguish
the Coulomb explosion mechanism from photothermal surface
evaporation. Here three temperatures were considered: electron
temperature Te , lattice temperature TL , and the temperature of the
medium surrounding the particle Tm .
Ce (Te )
dTe (t)
= −g (Te ) · (Te − TL ) + S (t)
dt
(7a)
CL (TL )
dTL (t)
= g (Te ) · (Te − TL ) − F
dt
(7b)
∂Tm (r, t)
1 ∂
Cm (Tm )
= 2
∂t
r ∂r
km r
2 ∂Tm
(r, t)
∂r
+F
(7c)
Table 1
Threshold laser power densities taken from available literature.
Target particle (diameter)
Pulse duration
Au sphere (19–47 nm)
Au sphere (25 ± 8 nm)
Au sphere (10–100 nm)
Au sphere (10–100 nm)
Au sphere (10–100 nm)
Au sphere (55 ± 7 nm)
Au sphere (55 ± 7 nm)
Au sphere (55 ± 7 nm)
Ag sphere (40–60 nm)
Au sphere (10 nm)
Au sphere (3.5 nm)
Au sphere (3.5 nm)
Au sphere (60 ± 8 nm)
Au sphere (60 ± 8 nm)
Au sphere (38 nm)
Au sphere (80 nm)
7 ns
30 ps
10 ns
10 ns
10 ns
4–6 ns
4–6 ns
4–6 ns
18 ps
10 ns
15 ps
20 ps
150 fs
150 fs
100 fs
∼6 ns
a
b
Excitation wavelength
532 nm
355 nm
355 nm
532 nm
1064 nm
266 nm
355 nm
532 nm
355 nm
355 nm
355 nm
532 nm
400 nm
532 nm
400 nm
355 nm
Threshold energya
Mechanismb
Threshold peak power
Year
Source
40–50 mJ cm−2
17 mJ cm−2
NA
NA
NA
20 mJ cm−2
24 mJ cm−2
33 mJ cm−2
NA
NA
NA
NA
7.3 mJ cm−2
3.6 mJ cm−2
9–12 mJ cm−2
120 mJ cm−2
PT
PT
PT
PT
PT
PT
PT
PT
CE
CE
CE
CE
CE
CE
NF
6–7 MW cm−2
560 MW cm−2
300 MW cm−2
150 MW cm−2
1999
2005
2009
2009
2009
2011
2011
2011
1998
2007
2010
2010
2011
2011
2006
2007
[90]
[91]
[92]
[92]
[92]
[100]
[100]
[100]
[94]
[96]
[99]
[99]
[70]
[70]
[106]
[118]
NA: not available.
PT: photothermal evaporation; CE: Coulomb explosion; NF: near-field ablation.
3.8 MW cm−2
4.5 MW cm−2
6.2 MW cm−2
27 GW cm−2
1.2 GW cm−2
0.5 GW cm−2
46 GW cm−2
23 GW cm−2
90 GW cm−2
20 MW cm−2
S. Hashimoto et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 13 (2012) 28–54
Equation (7a) describes the evolution of the electron temperature Te coupling to the lattice with the electron–phonon coupling
constant of g(Te ), and the laser irradiation term S(t). Equation (7b)
describes the dynamics of the lattice temperature TL that include
heat transfer across the interface between the particle and its surroundings, expressed by a term F. Finally, equation (7c) describes
the medium temperature on a spherical surface dependent both
on the time and distance from the particle center r. Here Ce (Te ),
CL (TL ), and Cm (Tm ) are the electron, lattice, and medium specific
heat capacities, and km is the thermal conductivity of medium. S(t)
and F are given by
S (t) =
F=
Cabs
(nm , R) · P (t)
Vp (TL )
h · AP
[TL − Tm (R)]
VP
points were important in estimating the temperature dynamics but
were not considered previously [69].
According to the investigation by Werner’s group [100], the
Coulomb explosion mechanism operates if the electron temperature exceeds a certain threshold. This is because, at high Te such as
5000 K, sufficiently energetic electrons in the tail of the Fermi distribution can overcome the work function of a metal and escape from
the surface. This process is known as thermionic electron emission
[97,98]. The probability of electrons thermionically emissible from
one atom can be calculated by
∞
nε =
(8)
(9)
Equation (8) shows that the energy absorption by laser irradiation is described by the absorption cross section Cabs (m2 )
multiplied by the laser power density, P (J m−2 ) and divided by the
particle volume Vp (m3 ). Most importantly, S(t) represents energy
truly absorbed by a NP irradiated with a laser energy density of P.
In equation (9), h (a typical value is 100 MW m−2 K−1 for 100-nm
Au NPs in water [89]) is the interface thermal conductance, Ap (m2 )
is the particle surface area, and Tm (R) is the maximum water temperature at the NP–water interface. Note that, in equations (8) and
(9), R denotes the particle radius.
The simulation model, which is based on a previous thermophysical calculation of Au nanorods on the excitation by
femtosecond-laser pulses [69], can be improved, especially for
ultrashort pulse heating of Au NPs by taking advantage of a previous
theoretical treatment [113]. An exhaustive description of the electron heat capacity and the electron–phonon coupling factor was
complied by applying the real electron density of states, Fermi distribution, and temperature-dependent chemical potential. It has
been revealed that the free-electron gas model (FEG) with its linear temperature-dependent electron heat capacity (Ce (Te ) = Te ),
assuming a parabolic electron density of states and a constant
electron–phonon coupling factor, is valid only at temperatures
below the electron temperature of 2500 K [113]. Above 2500 K, the
real electron heat capacity differs significantly from that of the FEG
model. For instance, at 10,000 K, the electron heat capacity of gold
is three times higher than that derived from the FEG model. Importantly, this high electron temperature can be easily attained via
femtosecond-laser excitation. For instance, Fig. 7a shows the simulation results of temperature evolution based on the FEG model.
Here exactly the same laser parameters were used as those used in
the rigorous model that leads to the computational result given in
Fig. 3 for 55-nm aqueous Au NP colloids.
A comparison of Figs. 3 and 7 may deserve attention. Significantly overestimated Te values were obtained in the initial stage
when the FEG model was used (Fig. 7a). However, the values of TL
at 60–100 ps delays in Fig. 7a are similar to those observed in Fig. 3,
suggesting that the estimation of TL is still acceptable for the calculation based on the FEG model. Fig. 7b gives the temperature profile
when the heat dissipation to the surrounding water is intentionally
neglected. Fig. 7b shows that heat losses can be neglected to obtain
TL in the femtosecond-pulse excitation with reasonable accuracy
because of the slow heating process of the surrounding water. This
explains why the heat dissipation was negligible in the temperature calculation of Au nanorods excited by a femtosecond laser [82].
The limited applicability of the FEG model was observed for the
electron–phonon coupling factor [113]. Werner’s group adopted
temperature-dependent density changes of the nanoparticle and
water and a Gaussian time profile of a laser pulse in S(t) [100]. These
37
EDOS (E) f (E, (Te ) , Te ) dE
(10)
ε
where EDOS(E) is the electron density of states, f is the Fermi distribution defined as f = (exp[(E − )/kB Te ] + 1)−1 dependent on the
chemical potential , and Te . The fission probability depends on
the ratio of the Coulomb expulsion term to the surface energy of
a particle: if the former exceeds the latter, charged droplets are
unstable to undergo explosion, which has been known since the
pioneering study by Rayleigh (Rayleigh instability) [114,115]. The
fragmentation temperature Te (frag), at which the electron temperature exceeds the threshold to emit a sufficient number of electrons
to cause instability, can be obtained. For instance, for a 60-nmdiameter Au NP, Te (frag) can be calculated as 7300 K in the liquid
state, and as 8200 K in the solid state. At 7300 K, the number of
electrons ejected is calculated as 600, and at 8200 K, it is 1500.
Obviously, the liquid NP is less stable than the solid because of the
reduced surface energy.
The temperature calculation in the nanosecond regime is
more simply performed because the electron–phonon coupling
is completed in the nanosecond-time domain. Thus, equation
(7a)–(7c) reduces to a set of equations given in the following (onetemperature model (OTM)) [116]
CL (TL )
dTL (t)
= −F + S(t)
dt
Cm (Tm )
∂Tm (r, t)
1 ∂
= 2
∂t
r ∂r
(11a)
km r 2
∂Tm (r, t)
∂r
+F
(11b)
Fig. 8 shows temperature versus time curves obtained by a simulation using equation (11a) and (11b) for the nanosecond-laser
excitation.
Two different laser power densities were employed in the simulation. The increase in particle temperature (or lattice temperature
TL ) takes place monotonously with time, passing through the mp
of bulk gold and reaching the bp of bulk gold, in both Fig. 8a and
b. At much higher laser power densities, however, a much rapid
temperature increase is noted, as shown in Fig. 8b. Note that even
with the simulation using the TTM given in equation (7a)–(7c), the
time profiles of Te and TL obtained are quite similar to those of the
OTM. This suggests that as long as nanosecond-laser pulse is used,
no mechanistic difference can be expected between the TTM and
OTM. The simulation also suggests that bubble formation can take
place during the nanosecond-laser pulse duration because of the
explosive evaporation of water, superheated well above the bp of
water up to the critical temperature (647 K). The formation of a bubble around a NP presumably induces the reduction in the refractive
index from 1.33 to 1.0. As a result, a significant reduction in the
optical absorption of Au NP can result, as depicted in Fig. 9. The
thermal insulation of NPs within the bubble is also expected.
Importantly, the absorption cross section (Cabs ), which determines the absorbed laser power in equation (8), largely depends
on the temperature-dependent optical properties of both the Au
NP and the surrounding medium. In the femtosecond-pulsed-laser
excitation regime, where the pulse duration is smaller than the
38
S. Hashimoto et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 13 (2012) 28–54
a
b
Fig. 7. Time-dependent temperature profiles of electron temperature Te (dotted curve), lattice temperature TL (solid curve), and the temperature of the medium surrounding
the particle Tm (dashed dotted curve), calculated for the 55-nm-diameter Au sphere by applying the FEG model (a) and the evaluation of Te and TL calculated when heat
dissipation to the surrounding water was neglected (b). Exactly the same laser parameters as those given in Fig. 3 were used.
electron–electron and electron–phonon relaxation time scales, Cabs
can be calculated with the optical properties at ambient temperature Cabs (nm (Tambient ), R). Here the pulse energy is totally absorbed
by the NP before the temperatures of the electrons, lattice, and surrounding medium increase significantly. On the other hand, in the
nanosecond-pulsed-laser excitation regime, Te , TL , and Tm increase
simultaneously during the Gaussian pulse-time profile. Thus, Cabs is
dependent on temperature, assuming the following very complex
form: Cabs (nm (Tm , p), R, TL , RB ). Here the medium refractive index
nm is a function of the temperature and pressure gradients around
a
298
b
the Au NP. If Tm exceeds the threshold of nanobubble formation, the
refractive index in the vicinity of the Au NP decreases to a minimum
value of 1.00, and the value of Cabs is significantly reduced inside the
expanding bubble with a radius of RB . Furthermore, lattice heating
induces the changes in the optical properties of the Au NPs, which
have been mentioned above (see Fig. 2). All these points make
the temperature calculations in the nanosecond-excitation regime
very difficult because of insufficient knowledge about the behavior of the nanobubble expansion and collapse. However, it should
be pointed out that without considering of this bubble effect, an
accurate estimation of the size-reduction threshold is not possible.
3.2.3. Interpretation of experimental size reduction based on
temperature simulation
The experimental study of size reduction can be demonstrated
by in situ absorption (extinction) spectroscopy assisted by transmission electron microscopy (TEM). Fig. 10a shows an example of
in situ spectral changes recorded for the size-reduction experiment
of the 60-nm-diameter Au NP excited by a femtosecond laser with
a wavelength at 400 nm. Fig. 10b shows the plots of experimental
A, the changes in LSPR band peak intensity, after 3,600,000 shots
(1 kHz, 60 min) as a function of the laser power density (scale on
the left) on excitation at 400 nm.
In Fig. 10b, a slightly decreased A occurs at a laser power density of 2.5 mJ cm−2 and A remains constant up to 5.0 mJ cm−2 .
When the laser power density increases to 7.3 mJ cm−2 , a greater
-15
x10
298
Fig. 8. Temperature versus time curves for lattice (particle) temperature TL
(solid curve) and maximum water temperature Tm at the NP–water interface
(dashed–dotted curve) for a 55-nm-diameter Au particle interacting with a 5ns laser pulse (FWHM of the Gaussian time profile) at 355 nm, 28 mJ cm−2
(Pmax = 5.26 × 106 W cm−2 ) (a) and 52 J cm−2 (Pmax = 9.77 × 109 W cm−2 ) (b). See the
caption of Fig. 3 for symbols Tbp , Tmp , Tcp . (Modified from ref. 100. Reproduced with
permission from the American Chemical Society.)
Fig. 9. Absorption cross section curves as a function of wavelength for a 55-nmdiameter Au NP in water of various phases: liquid (n = 1.33), critical state (n = 1.07),
and vapor (n = 1.00). (Reproduced from ref. [100]. Reproduced with permission from
the American Chemical Society.)
S. Hashimoto et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 13 (2012) 28–54
b
1.5
0 min
15 min
40 min
60 min
A
1
Max. temperature / K
extinction
a
39
0.5
298
0
400
500
600
700
800
wavelength /nm
Fig. 10. (a) Typical time sequence of in situ spectra for 60-nm Au NPs in aqueous solution during femtosecond-pulsed-laser irradiation at 400 nm, and a laser power density
of 19.4 mJ cm−2 . The spectra were recorded at 0, 15, 40, and 60 min. The repetition rate was 100 Hz. (b) Plots of experimental A, the change in the LSPR band peak intensity
after 3,600,000 shots (1 kHz, 60 min) as a function of laser power density (scale on the left) on excitation at 400 nm by a femtosecond laser (150 fs). Calculated laser power
density dependent maximum temperature evolution of Te (dashed red line) and TL (solid black line) for a 60-nm aqueous Au NP (scale on the right) is also included. The
vertical dashed-dotted lines (q1 = 6.0, q2 = 7.4 mJ cm−2 ) indicate the range of the calculated Coulomb explosion threshold for liquid gold nanodroplets. The dashed line shows
the reshaping threshold at 700 K. (Reproduced from [70]. Reproduced with permission from the American Chemical Society.)
Fig. 11. TEM images of 60-nm Au NPs after 60 min of femtosecond-laser irradiation at 100 Hz at an excitation wavelength of 400 nm, (a): 0 mJ cm−2 , (60 ± 8) nm; (b):
3.7 mJ cm−2 , (55 ± 5) nm; (c): 7.6 mJ cm−2 ; (57 ± 13) and (2.5 ± 1.3) nm; (d): 12.1 mJ cm−2 , (56 ± 13) and (3.2 ± 1.2) nm, (e): 19.4 mJ cm−2 , (54 ± 12) and (3.0 ± 1.3) nm. (Modified
from ref. [70]. Reproduced with permission from the American Chemical Society.)
a
Max. temperature / K
b
Max. temperature / K
decrease in A occurs and continues at higher power densities.
Thus, the threshold energy of size reduction was assigned to
7.3 mJ cm−2 [70]. The assignment was consistent with TEM images
given in Fig. 11, because the fragmentation started approximately
at the same laser power density.
This threshold laser energy cannot be explained by the photothermal mechanism, because the lattice temperature is still far
below the bp of gold at this stage (Fig. 10b, TL ≈ 1300 K). However, at
this laser power density, Te can exceed the fragmentation threshold
of 7300 K for a liquid gold droplet. Thus, it is reasonable to consider
that the Coulomb explosion mechanism is applicable.
On the other hand, if we look at the A versus the laser power
density profile for a nanosecond-laser excitation, a completely different picture emerges. Fig. 12 is such an example. Since Te = TL for
the nanosecond-laser excitation, extremely high Te that exceeds
Te (frag) cannot occur. Instead, the maximum TL increases steadily
with increasing laser power densities, as shown in Fig. 12. Thus, the
size-reduction mechanism is described exclusively by the evaporation mechanism that occurs when the particle temperature exceeds
the bp.
As stated above, in nanosecond-laser excitation, heat loss to the
surrounding medium and the optical effect of bubble formation
during the excitation are important. In Fig. 12, the particle temperature curve was constructed considering these two effects. If
the heat loss and bubble effect are totally neglected, a considerably
different temperature curve can be drawn, especially for the excitation at 532 nm, where the LSPR peak position is located nearby.
Fig. 13 shows such an example.
Fig. 12. Laser power density dependent evolution of maximum TL (solid line) for
a 55-nm aqueous gold NP (scale on the right side) together with the experimental
plots of A for 36,000 shots versus the laser power density (scale on the left side) on
excitation at 532 nm (a) and 266 nm (b). Vertical lines represent the experimental
thresholds of melting and evaporation. (Modified from ref. [100]. Reproduced with
permission from the American Chemical Society.)
40
S. Hashimoto et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 13 (2012) 28–54
a
max. temperature /K
b
max. temperature /K
Fig. 13. Temperature curve calculation of maximum TL (solid line) for a 55-nm
aqueous Au NP (scale on the right side) when the heat loss and the optical effect
of bubble were neglected, together with the experimental plots of A for 36,000
shots versus the laser power density (scale on the left side) on excitation at 532 nm
(a) and 266 nm (b).
Although the experimental size-reduction threshold, i.e., the
evaporation threshold for the excitation at 532 nm, is determined
as 33 mJ cm−2 , the computationally obtained bp is 10 mJ cm−2
(Fig. 12a). This large discrepancy is significantly reduced in
Fig. 12a in which the computational particle boiling threshold is
28 mJ cm−2 . Inspection of Fig. 12b reveals that this discrepancy
is small for the excitation at 266 nm. This is because the optical
effect of bubble formation is dependent on the excitation wavelength and is not significant at 266 nm (see Fig. 9). It is important
to note that previous studies of laser-induced size reduction completely ignored the two effects, which leads to the overestimation
of temperature increase [90–99].
The combination of the experimental laser irradiation and
computational temperature studies of Au NPs revealed that the
nanosecond-laser-induced size reduction can be represented by
Scheme 5.
A possible scenario is as follows. During the nanosecond-pulsedlaser heating of Au NPs, thermal energy is transferred to the
surrounding water and its temperature, Tm , increases steadily.
As the spinodal temperature of 573 K [73] is attained, explosive
evaporation of the water occurs and a “hot” high-pressure bubble
expands rapidly, which depresses Cabs through a refractive-index
reduction in the LSPR region. If the laser intensity is sufficiently
high to heat up the thermally insulated NPs to the boiling point,
surface evaporation, and thus, the size reduction of the original Au
NPs results. The generated fragments may form aggregates because
of low permittivity within the bubble. Finally, after the nanobubble collapses and the NP cools down to room temperature, the
coexistence of size-reduced original NPs and small “snake-like” or
“nut-like” structures can be observed in TEM images.
So far only one study has reported the failure of the
observation of single-shot femtosecond-laser-induced fragmentation. Warshavski et al. [117] used both 800-nm (50 fs,
150 mJ cm−2 , 3 × 1012 W cm−2 ) and 550-nm (90 fs, 90 mJ cm−2 ,
1.8 × 1012 W cm−2 ) laser pulses for the excitation of 16, 23, and
45 nm diameters of Au NPs in aqueous solution. They could not
identify any statistically significant decrease in the average particle
size by TEM observation. They ascribed the effect of femtosecond
pulses as causing no significant particle size reduction. However,
the interpretation of an ensemble experiment, in which only a portion of the particles that are exposed to the peak intensity of a laser
pulse undergoes complete fragmentation, requires a certain precaution. The particles exposed to the wings of a spatial Gaussian
beam may be reshaped rather than fragmented because of insufficient energy absorption.
3.2.4. Time scale of fragmentation
To reveal the time scale of size reduction/fragmentation,
transient-absorption spectroscopy was employed by Werner
et al. [70]. Previously, femtosecond pump-probe spectroscopy has
been applied to elucidate the electron dynamics of plasmonic
Scheme 5. Schematic illustration of the size reduction of aqueous gold NP colloids by nanosecond-pulsed-laser irradiation.
S. Hashimoto et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 13 (2012) 28–54
41
Fig. 14. Transient-absorption spectra of 30-nm Au NPs in water as a function of the time delay between the pump- and probe-(white light) laser beams. The excitation
wavelengths are (a) 600 nm (intraband) and (b) 380 nm (interband). The insets show the kinetics of the plasmon band at 530 nm. (Reproduced from ref. [66]. Reproduced
with permission from the American Chemical Society.)
NPs [65–67]. Fig. 14 shows an example of transient-absorption
spectral changes of Au NPs in aqueous solution. Negative absorption or “bleaching” signals are indicative of ultrafast electron
relaxation (cooling) processes (electron–electron relaxation and
electron–phonon relaxation processes), mentioned above.
The bleaching signal occurring from initial femtoseconds to
several hundred picoseconds may interfere with transient signals associated with laser-induced fragmentation because the size
reduction can be observed by the reduction in the LSPR band. However, it is interesting to note that an isosbestic point appears at
≈490 nm in the transient-bleaching signals, as shown in Fig. 14. By
exciting 60-nm-diameter Au NPs at 400 nm and probing at 490 nm,
the LSPR bleaching signals ascribable to the fragmentation of Au
NPs were observed (Fig. 15).
At a laser power density of 3.7 mJ cm−2 which is below the
fragmentation threshold, only a slightly reduced signal ascribed to
a surface melting was observed. At the fragmentation threshold
(6.1 mJ cm−2 ) and above (17.2 mJ cm−2 ), greater bleaching signals
increasing with laser power density were observed and this was
ascribed to fragmentation signal. Inspection of Fig. 15 reveals that
fragmentation takes place not instantaneously but over 100 ps. This
fragmentation scheme was proposed by assuming the Coulomb
explosion mechanism as in the following Scheme 6.
On excitation with a laser power density above 6–7 mJ cm−2 ,
high Te of greater than 7300 K can easily be achieved to induce
thermionic electron emission. At the same time, TL increases to
1337 K, causing a phase transformation from solid to liquid within
Fig. 15. Femtosecond-transient-bleaching signals at various power densities versus
time for 60-nm-diameter Au NPs in aqueous solution measured at (490 ± 5) nm.
(Reproduced from ref. [70]. Reproduced with permission from American Chemical
Society.)
3–10 ps. This phase transformation greatly decreases the surface
tension (surface energy) of the Au NP. When the repulsion energy
caused by the critical charge of the liquid Au NP exceeds the
Rayleigh instability threshold, the liquid droplet splits into many
smaller nanodroplets, that separate into individual clusters in the
time course of 100 ps. This is a picture obtained by the transientabsorption study given in Fig. 15.
Scheme 6. Schematic illustration representing the femtosecond-laser-induced fragmentation process. (Reproduced from ref. [70]. Reproduced with permission from the
American Chemical Society.)
42
S. Hashimoto et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 13 (2012) 28–54
By analyzing the time-resolved Bragg reflection intensity on the
irradiation of a 100-fs pulsed-laser ( = 400 nm), Plech et al. demonstrated the lattice expansion and melting of 38-nm-diameter Au
NPs in water [106]. They reported that the surface ablation of Au
NPs occurs at 9–12 mJ cm−2 before the complete melting of the particle that occurs at a power density of 15 mJ cm−2 . A fragmentation
model was proposed that assumes the spallation of Au ions from
the surface of the Au NPs within 100 ps of excitation caused by the
near-field enhancement of the incident laser light.
3.2.5. Single-particle approach
As we mentioned above, ensemble measurements suffer a drawback in that they only observe an average figure of various events
caused by various laser power densities for various particles sizes.
Thus a single-particle approach is favored, because it could give
an exact and precise picture of a laser-induced event. Along this
line, Ito et al. [118,119] conducted the laser-ablation of single 80nm-diameter Au NPs submerged in water on a glass substrate. The
laser manipulation technique [120–122] was used to bring a NP to
a desired position on the substrate. The laser irradiation was conducted by irradiating a single-shot or multiple shots of a Nd:YAG
laser pulse (355 nm, 6 ns). Atomic force microscopy (AMF) images
were recorded before and after the irradiation. The fragmentation threshold of 120 mJ cm−2 was determined by the observation
of ablation-induced fragments. The threshold value is appreciably
higher than that of solution measurements because of substrate
effect (see Fig. 12). It was also found that the individual ablation of
Au NPs on glass substrates in a solution results in the efficient adhesion of their fragments onto the substrates. This can be used for the
formation of patterned NPs on a substrate. For swift and convenient
analysis of particles without the aid of AFM, single-particle spectroscopy measurements can be performed to monitor the ablation
event by applying dark-field microscopy [59].
3.3. Heat transfer from gold nanoparticles to the surrounding
medium
Hu and Hartland [123] investigated the cooling dynamics of Au
NPs with various diameters dispersed in water by femtosecond
pump-probe spectroscopy. From their experiment, they concluded
that the probe signal versus time profile could be conveniently
fitted by applying a stretched exponential function, given by
ˇ t
F (t) = A · exp −
(12)
Here is the cooling time constant, t is the delay time, and ˇ
is the stretching parameter. By fitting the experimental results,
they found an almost quadratic dependence of the cooling times
on the diameter of the particles, whereas the stretching parameter remained practically constant at 0.7. Fig. 16 shows the Au NP
size-dependent cooling time constants.
Hu and Hartland gave an empirical equation for calculating the
NP temperature at low temperatures;
T (t) = (Ti − T0 ) · exp −
t
· R2
0.159
0.7 + T0
(13)
where Ti is the initial temperature at thermal equilibrium and T0 is
the temperature of the surrounding water.
Subsequently, Plech et al. [89] measured the cooling dynamics by observing the lattice expansion and contraction of Au
NPs dispersed in water. They analytically solved a set of two
differential heat diffusion equations in the Laplace domain and
additionally introduced a thermal heat conductance, h, at the
NP–water interface. By fitting their analytical results with the
experimentally obtained data, they estimated a heat conductance
Fig. 16. Characteristic time constant for energy dissipation determined using equation (12) versus diameter: () experimental data; (–) calculated temperature versus
time profiles. The dashed line shows a fit to the data assuming a parabolic dependence of on diameter (i.e., ∝ R2 ). (Reproduced from ref. [123]. Reproduced with
permission from the American Chemical Society.)
of (105 ± 15) MW m−2 K−1 . It was considered that the NP–water
interface builds a layer of a few nanometers surrounding the NP
surface with an average temperature defined by Tm (R). For a sphere
with radius R, the surface area AP , and the volume VP , are defined
as AP = 4R2 and VP = 4/3R3 . Thus the heat transfer term at the
NP–surface–water boundary, equation (9), reduces to
F=
3h
[TL − Tm (R)] .
R (TL )
(14)
Equation (14) shows that the heat loss through the NP–water
interface is proportional to the temperature gradient between the
lattice and water. Furthermore, F is inversely proportional to the NP
radius, which consequently leads to an increase in the heat transfer
by reducing the NP size. In other words, a strong thermal nonequilibrium exists between TL and Tm , and smaller NPs have greater
heat loss. Depending on the distance from the NP–water interface,
the surrounding water medium gives a temperature distribution
according to equations (14) and (7c) (or (11b)). Fig. 17 shows an
example of the calculated spatial water temperature curves for a
55-nm Au NP excited with different pulse durations, 5 ns and 150 fs,
at three different time delays.
Inspection of Fig. 17 reveals that the nanosecond-laser heating of
Au NPs leads to a greater heat dissipation to the surrounding water
than the femtosecond-laser heating. The lattice and water temperatures at the NP surface–water interface are in quasi-equilibrium for
the nanosecond case. In contrast, in the femtosecond-laser heating,
the effect of heat conductance is clearly visible. At a short delay of
100 ps, a discontinuous decrease is observable in the temperature
across the NP–water interface because of the effect of the thermal
boundary conductance h: a thermal non-equilibrium of TL and Tm
is established. By increasing the delay time, the difference between
TL and Tm is reduced and vanishes after 600 ps. At this time delay,
a thermal equilibrium is attained caused by the cooling of the NP
lattice temperature occurring simultaneously with the heating of
the surrounding water. Thus, the results show that the finite thermal interface conductance is an important factor for controlling
the cooling of metal NPs in condensed media. Molecular dynamic
simulations of the heat transfer of Au NPs in different liquids have
shown temperature distributions similar to those given in Fig. 17
[124]. The heat conductance of Ag NPs embedded in a glass matrix
has been estimated by Juve et al. [125].
An important consequence of the heat transfer to the surrounding liquid of laser-heated metal NPs is the formation of
S. Hashimoto et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 13 (2012) 28–54
43
Fig. 17. Calculated temperature profiles of water surrounding a 55-nm-diameter Au NP excited with a laser power density of 2.5 mJ cm−2 at (a) 355 nm (pulse duration: 5 ns)
and (b) 400 nm (pulse duration: 150 fs) as a function of the radial distance r from the particle center. The time delays are (a) 7.5 ns (black curve), 10 ns (red curve), and 15 ns
(blue curve); (b) 100 ps (black curve), 200 ps (red curve), and 600 ps (blue curve). The starting point of the simulation was −2 P from the maximum of the Gaussian time
profile. The temperature inside the particle is assumed to be uniform. Temperatures, Tcp and Tcav , represent the temperatures of water at the critical point (647 K) and the
bubble formation threshold (573 K), which considers the spinodal effect of water.
vapor nanobubbles. Plech and co-workers [42,126] experimentally
investigated the formation of vapor bubbles by probing the environmental water and its structural change with X-rays after the
explosive phase transformation. They estimated that the vapor
nanobubble with an initial high pressure expands up to a maximum
radius, where the inner pressure is reduced to an equivalent value
to the outer water pressure, and then the bubble starts to collapse.
The maximum radius greatly depends on the laser power. After the
first cycle, the surrounding liquids are heated up again to the critical
temperature through heat transfer from the NP surface and another
bubble formation occurs. This process is repeated in a sequential
manner with a decrease in the maximum bubble radii until the
metal NP is cooled down far below the cavitation temperature.
An example of measured bubble dynamics by X-ray pump-probe
spectroscopy together with the calculated bubble radii and inner
pressures by applying the Rayleigh–Plesset equation is given in
Fig. 18.
Here the maximum bubble radius is reached after 250 ps, at
which time the pressure within the bubble decreased to an equivalent value of the outermost water pressure (1 atm). The first
maximum in pressure at 650 ps denotes the collapse of the first
bubble. The following calculated modulations in pressure and size
are only expected for oscillatory bubble motion.
Subsequently, Plech and co-workers [43] conducted the excitation of Au NPs dispersed in water by a femtosecond-pulsed-laser
and observed their lattice contraction during the cooling process.
At a certain laser power density and delay time, the NP lattice remained constant without undergoing contraction, where no
Fig. 18. Bubble radius (solid curve) and pressure (dashed curve) transients of the
water vapor inside the bubble as calculated from the Rayleigh–Plesset equation
together with the measured radii. (Reprinted with permission from ref. [42]. Copyright 2005, American Institute of Physics.)
efficient heat transfer occurred. They assigned this laser power
density as the threshold power density of the formation of a
nanobubble that thermally insulates the Au NPs. Fig. 19 gives the
relative lattice expansion dependent on the laser power density for
two Au NP sizes.
Furthermore, they estimated a bubble formation temperature
that is slightly dependent on the Au NP size but is approximately
equal to the critical point of water (Tcp = 647 K).
More recently, Plech’s group [127] conducted nanosecondtime-resolved spectroscopy experiments on 60-nm and 80-nmdiameter Au NPs dispersed in water to observe the bubble
dynamics. They excited the suspension by single 10-ns pulses
at 532 nm or 355 nm to measure the extinction signals with
Fig. 19. Lattice expansion of Au NPs of 52- and 94-nm diameters as determined from
the peak shift of the (1 1 1) powder reflection at 100 ps (circles) and 1 ns (crosses)
delays together with a calculation of the thermal expansion (lines; dashed line without rescaling). Above 155 J m−2 (15.5 mJ cm−2 ), no powder reflection at the 100 ps
delay is detectable, indicating particle melting. (Reprinted with permission from ref.
[43]. Copyright 2006, American Institute of Physics.)
44
S. Hashimoto et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 13 (2012) 28–54
Fig. 20. Transient-extinction change of an 80-nm Au NP suspension by four collinear multicolor laser probe beams excited with a single 10-ns laser pulse at 532 nm at a
power density of (A) 300 J m−2 (30 mJ cm−2 ) and (B) 450 J m−2 (45 mJ cm−2 ). (Reproduced from ref. [127]. Reproduced with permission from IOP Publishing.)
nanosecond-time resolution at several wavelengths. They observed
an oscillating signal of one period from a negative signal turning to
a positive one, which is given in Fig. 20.
Plech’s group discussed the detected extinction signals on the
bases of the calculated Mie extinction spectra of a Au NP and
expanding vapor bubble in water at different temperatures. They
concluded that the extinction signal at the threshold laser power
density of bubble formation first decreases, because of a reduced
extinction of Au NPs caused by a refractive-index change from
water to vapor (from 1.33 to 1, see Fig. 9) and a negligibly small
extinction signal of small vapor bubbles. At longer time delays,
the vapor bubbles grow, leading to an extinction (scattering) signal
higher than that of the Au NP extinction at ambient temperature.
The right side of Fig. 21 shows an extinction change
∈(˚)/∈ (0)|t=5ns as a function of the laser power density. A calculation of the extinction as a function of the bubble size is shown on the
left. The interpretation assumes that the extinction first decreases
on increasing the power density (increasing maximum bubble size)
because of the decoupling of the LSPR from the dielectric surrounding. At a larger laser power density, the bubble increases in size and
enhances scattering as a positive contribution to the extinction. It
should be mentioned that Plech and co-workers applied rather high
laser power densities, well above the size-reduction threshold of Au
NPs dispersed in water [90,93]. Size reduction and reshaping may
lead to a permanent extinction reduction and blue shift of the plasmon band, and thus should be considered in the interpretation of
the signal observed here.
Lapotko’s group [45,128–130] carried out scattering spectroscopy on a single Au NP placed on a glass substrate surrounded
by water. They excited the NP with a single laser pulse with
both nanosecond- and subnanosecond-time scale durations. They
detected the time-dependent scattering and transmission signals using a CW or pulsed-laser for analyzing light. They found
that a much higher laser power density (100 mJ cm−2 for a 250nm-diameter Au) at an excitation laser wavelength of 532 nm
is necessary for the threshold of bubble formation than that
expected theoretically by employing a simple light-to-heat conversion model. At the threshold of bubble formation, a minimum
lifetime of ∼9 ns was observed, which was independent of the Au
NP diameter. This discontinuity (threshold behavior) could not be
detected by replacing the Au NPs with absorbing molecules dissolved in water. In this case, the shortest bubble lifetime measured
by a photodetector (transmittance) signal at the vapor bubble generation threshold was ∼1.6 ns, and the lifetime increased gradually
with increasing laser power densities with a very low threshold. Because of the much higher bubble generation threshold and
discontinuity in the minimum lifetime, Lapotko’s group considered the phenomenon a new complex nanosystem, i.e., plasmonic
nanobubble (PNB). It should be pointed out that they did not consider the possible effects of glass substrate on cooling and bubble
nucleation. Fig. 22 shows the laser power density dependent bubble
generation probability and lifetime of a 90-nm Au NP.
Lapotko’s group discussed the increased threshold for bubble
formation on the basis of the curvature effect. According to the
Young–Laplace equation, a smaller curvature radius of water yields
a higher vapor pressure. This additional pressure has to be overcome before the bubble, which is generated by a laser-heated Au NP,
can expand into the surrounding water. In addition, they detected
increased scattering or extinction signals by increasing the laser
power density. Thus, the lifetime and maximum scattering was
directly correlated with the bubble size.
The experimental data of Plech [127] and Lapotko [129] contradict to each other. For instance, one significant difference is the
threshold laser power density of bubble formation. Lapotko’s group
postulated much higher laser power densities necessary for bubble initiation than the theoretically estimated ones. On the other
hand, Plech’s experimental data agreed well with the thermophysical calculations and the spinodal effect of water. Also, Lapotko did
not observe any oscillation-like signals.
Neumann and Brinkmann [44,131,132] conducted experimental investigations on the nucleation dynamics of bubbles in
water by the 532-nm pulsed-laser excitation (duration: 12 ns) of
S. Hashimoto et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 13 (2012) 28–54
45
Fig. 21. Left: calculation of the extinction cross section of an Au NP with a radius of 30 nm (60 nm diameter) as a function of a surrounding vapor bubble of variable size.
Vertical lines denote the particle–vapor interface. Right: measured extinction change at 532 nm through a 60-nm-diameter Au NP suspension in the flowing capillary at a
variable laser power density of 355-nm pulses. (Reproduced from ref. [127]. Reproduced with permission from IOP Publishing.)
microabsorbers, melanosome. They found a rather low bubble
threshold at a temperature of 409 K. They explained the observation by the spinodal effect of superheated water, where the density
of water becomes dramatically low (spinodal density) and explosively converts into vapor. In other words, the average distance
between the water molecules reaches a critical distance because
of the disruption of hydrogen bondings. Furthermore, Neumann’s
group obtained the bubble radii and expansion velocities from
time-resolved microscopy dependent on the laser power density.
The velocities ranged from 10 to 85 m s−1 for laser power densities that were 1.5–8.5 times greater than the threshold power
density. On the other hand, the shock wave velocity, generated by
pulse-laser-heated NPs, is approximately the velocity of sound.
The mechanisms of laser-induced bubble formations assisted
by nano-/micro-absorbers have not been fully understood. Clear
explanations for the experimental results still are needed.
3.4. Laser-induced agglomeration and fusion
Fig. 22. (A) Time responses for single 80-nm Au NPs (a–c) and a molecular solution of Trypan Blue (d–f) obtained with a single pump-laser pulse (vertical dashed
line): (a) subthreshold fluence representing medium heating, (b) near-threshold
power density of bubble formation, (c) over-threshold power density showing bubble dynamics, (d) dye absorption, (e) bubble formation power density, (f) bubble
dynamics. Two arrows in f show the deviation of the after bubble signal from the
baseline caused by residual heating of the surrounding water. (B) Dependence of the
PNB lifetime (TPNB , red) and the generation probability (black) around single 90-nm
Au spheres on the power density of a single pump-laser pulse (0.5 ns, 532 nm); vertical line separates the types of corresponding time responses (A), and the horizontal
line separates the optical scattering effect. (Reproduced from ref. [129]. Reproduced
with permission from the American Chemical Society.)
Instead of size reduction or fragmentation, the pulsed-laserinduced growth of Au NPs has been observed [133–144]. Kamat and
co-workers [133] observed that thionicotinamide-capped Au NPs
(particle diameter: 15–20 nm) dispersed in water undergo fusion
and fragmentation on laser-pulse excitation (532 nm, ∼18 ps,
1.5 mJ pulse−1 ) for 1 min. The resultant particles had grown in size
(∼100 nm) but were well separated. After 30 min of continuous
laser irradiation, they observed fragmentation of these particles,
producing small particles (diameter: 5 nm). They attributed this
phenomenon to the melting (fusion) of aggregates to form larger
spherical particles because surface-modified Au NPs exist as aggregates. On the other hand, similar laser-induced fusion was not seen
in uncapped-Au NPs.
Niidome et al. [136] observe that the irradiation of a
pulsed-1064-nm laser light (6–8 ns, 875 mJ cm−2 , 10 Hz) to
dodecanethiol-passivated Au NPs (3.2 ± 0.92 nm) in cyclohexane
resulted in the enormous growth of particles. The size increased
up to ∼200 nm with spherical shape. In this case, photofragmentation was unlikely, because the smallest particles coexisting were
∼10 nm. Removal and/or photodecomposition of the adsorbed
dodecanethiol molecules from the particle seems to be indispensable to the fusion of particles. The same group [138] reported
46
S. Hashimoto et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 13 (2012) 28–54
a
b
1.5
1.5
0s
30 s
2 min
4 min
10 min
60 min
1
extinction
extinction
1
0s
30 s
60 s
2 min
4 min
7 min
10 min
25 min
60 min
0.5
0.5
0
0
400
500
600
700
800
wavelength / nm
400
500
600
700
800
wavelength / nm
Fig. 23. Laser-induced agglomeration of Au NPs in an aqueous solution by the irradiation of a 532-nm pulsed-laser light (1.1 J cm−2 pulse−1 , 10 Hz) in a Ar-saturated solution
(a), and O2 -saturated solution (b). Au NPs were prepared by the laser-ablation in pure water on irradiation of a 1064-nm pulsed-laser light. (Reproduced from ref. [139].
Reproduced with permission from the American Chemical Society.)
that large and highly spherical Au NPs (diameter: 20–30 nm) can
be prepared by a pulsed-laser irradiation (532 nm, 6–8 ns, 10 Hz)
of polycation-Au NP (diameter: 10 nm) complexes in an aqueous solution. In this case, the complex of negatively charged
Au NPs and cationic poly(diallyldimethylammonium chloride),
PDDA was formed by electrostatic interaction. Later, Koshizaki
and co-workers exploited this technique for the preparation of
submicron-sized spherical particles with a homogeneous size distribution [140–143]. Large spherical particles were also prepared
by Pocovi-Martinez et al. [144]. Thiol-passivated Au NPs (3–4 nm
diameter) irradiated in chloroform at 266 nm (10 ns, 12 mJ pulse−1 )
led to a rapid increase (minutes time scale) in the size of NPs, producing large spherical particles (∼100 nm).
Werner et al. [139] observed the pulsed-laser-induced agglomeration of Au NPs leading to immediate flocculation in pure water.
Notably, they employed uncapped negatively charged Au particles
with a broad size distribution (10–100 nm diameter) prepared by
the laser-ablation technique [145]. The excitation was provided by
a 532-nm nanosecond-laser pulse (laser power density: 1.1 J cm−2 ).
The laser-induced particle growth process was monitored by the
extinction growth in the 700–800 nm region at the expense of
LSPR band depression. In this case, the size growth occurs simultaneously with size reduction because of the high-intensity laser
irradiation. This is the main cause of the observed LSPR band
depression. Fig. 23 shows the results.
The growth was seriously affected by the irradiation atmosphere: for instance, the retardation of growth was observed in
a Ar-saturated solution while the growth rate was enhanced in
a O2 -saturated solution (see the difference between (a) and (b)).
The mechanism of agglomeration was ascribed to the combination
reaction of cationic fragments generated by the laser-ablation of
Au NPs with intact Au NPs with an original negative surface charge
to produce large neutral species susceptible to further aggregation.
Thus, in water, the role played by the cationic species produced by
ablation-induced ionization is very important. In addition, species
may be generated, possibly by thermal processes such as melting
and vaporization.
Therefore, laser-induced melting, removal of stabilizing ligands, and charge neutralization are mainly ascribed to coalescence
(fusion) and agglomeration of particles for pulsed-laser irradiation.
Interestingly, it has been known for years that CW laser or lamp
irradiations can cause the coagulation of Au and Ag NPs [146–152].
This “photo-induced coagulation” proceeds for over many hours
(e.g. 30 h) for uncapped-Au NPs by the irradiation of UV light from
a high-pressure mercury lamp or a visible (488 and 514 nm) CW
laser beam in organic solvents such as 2-propanol and ethanol
[146–150]. Similar light-induced acceleration of aggregation for
aqueous Au NPs stabilized by citrate was observed by exciting the
LSPR band by a 514-nm CW laser [151]. According to Kimura, photoenhanced van der Waals attractive force on the excitation of the
LSPR band is responsible for the observed phenomena [148,149].
On this point, recent studies that apply an optical-trapping technique deserve comment [153–155]. Yoshikawa et al. demonstrated
that the optical trapping by a 1064-nm CW laser can control the
assembly of Au NPs (40 nm diameter) reversibly in water [153]. By
the action of optical force, the LSPR extinction band of assembled
NPs (600–700 nm) was observed to increase with increased laser
intensity owing to a conformational change in the Au NP assembly.
Zhang and co-workers [155] observed that the optical trapping of
micron-sized Au NP aggregates or 40-nm-diameter citrate-capped
Au NPs with a CW laser at 532 nm induced irreversible agglomeration. The observed trapping was explained by optical force, in
the first place, but they also attributed an additional cause to laser
heating.
4. Application of laser-induced response of metallic
nanoparticles
4.1. Surface modification and treatment of substrate
Interaction of lasers with plamonic NPs allows nanofabrication on various substrates on which the particles are placed
or assembled. Ko and co-workers reported the fabrication of
nanometer-sized craters on a polyimide film self-assembled with
Au NPs when exposed to a 532-nm nanosecond-laser light through
an objective lens (3–5 ns pulse width, 30–300 mJ cm−2 power
density) [156]. Tsuboi and co-workers observed the nanohole
(d < 100 nm) formation on a Au NP–polymer hybrid film deposited
on glass substrates and irradiated by a single 532-nm (∼8 ns) laser
shot [157]. These two studies attributed the origin of nanoholes to
a heat-mode fabrication, caused by the heat transfer from the NPs.
The heat mode of nanofabrication represents the material modification owing to the damage caused by melting and the explosive
evaporation of Au atoms and clusters. The nanoholes were generated through the rapid overheating of strongly absorbing NPs
S. Hashimoto et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 13 (2012) 28–54
47
Fig. 24. (a) SEM image of a borosilicate glass surface assembled with 47 ± 8 nm Au NPs with a surface coverage of 30% and exposed to a single-shot of a 532-nm pulsed-laser
light (power density: 360 mJ cm−2 pulse−1 ). The left side of the image shows the glass surface fully covered with small NPs resulting from the laser splitting of the original
particles. The right side of the image displays a bare surface from which the particles were removed after laser irradiation. Here many dark spots with a slightly bright outer
ring are formed on the bare surface. For comparison, the inset depicts the SEM image of the glass surface before the irradiation. (b) AFM topographical images of sample
surfaces subjected to a single-shot laser irradiation at a laser power density of 470 mJ cm−2 . Dark spots represent pitted areas corresponding to the craters. The average crater
depth is 4–6 nm. The AFM images were acquired for the surfaces exposed to the laser irradiation and in the absence of Pt–Pd coating. The fragments of Au particles were
removed before the measurement. (Reproduced from ref. [116]. Reproduced with permission from the American Chemical Society.)
during a short laser-pulse duration when the influence of heat
diffusion is minimal.
Hashimoto et al. [116,158] observed the formation of
nanocraters on a borosilicate glass surface assembled with
40-nm-diameter Au NPs when irradiated with a single-shot
nanosecond-pulsed-laser with a wavelength of 532 nm (Fig. 24).
Laser power densities of two orders of magnitude smaller than
those of causing the breakdown of the glass were applied for the
fabrication of the craters.
The average diameter of the craters was ∼20 nm with a depth
of less than 10 nm for the single-shot experiment: however, after
repeated irradiations, the size of the craters expanded to a 100 nm
scale [158]. The crater formation was simultaneous with the laserinduced splitting of Au nanoparticles to generate smaller particles
of 15 nm diameter. The number density of the craters increased
sharply, depending on the laser power density. The onset of the
increase occurred at ∼160–170 mJ cm−2 pulse−1 , and reached a
value of two times the number density of original Au particles
(150 ± 10 particles·␮m−2 ) with a weak tendency to level off at
high power densities (Fig. 25a). The onset of the crater formation coincides with the splitting of Au NPs because of increase in
the temperature above the boiling point (∼3100 K) of gold resulting from the absorption of the laser energy. Thus, the explosive
b
450
thermoacoustic stress / GPa
number density / m-2
a
300
150
0
evaporation of gold nanoparticles is postulated to play a crucial
role for the modification observed here. This assumption gained
a support from the estimation of laser power density dependent
thermoacoustic pressures due to the sudden evaporation of gold
(Fig. 25b).
This finding may represent a new application of the laserablation/fragmentation of nanoparticles to material processing
based on the photothermal process. Very recently, Tsuboi and coworkers have demonstrated that this laser-Au NP technique is
applicable to the formation of pores on the shell wall of spherical hollow silica microparticles embedded with Au NPs [159]. They
prepared pores of tunable sizes from 17 to 56 nm by varying the
size of original Au NPs from 6 to 39 nm.
In the meantime, Obara and co-workers observed the formation of nanoholes on the surface of a silicon substrate placed with
200-nm-diameter Au spheres by irradiating a single femtosecondlaser pulse (800 nm) with an intensity of less than the ablation
threshold of silicon [160–162]. Leiderer and coworkers observed
the ablation pattern for regular arrays of gold triangles with a side
length of 450 nm and a thickness of 25 nm when these structures
were illuminated with a 150-fs laser pulse (800 nm, 10 mJ pulse−1 )
[163,164]. Holes as small as 5 nm were fabricated at two corners of each triangle on a silicon substrate. Heltzel et al. observed
0
100
200
300
400
-2
-1
fluence / mJ cm pulse
500
60
40
20
0
0
100
200
300
400
500
fluence / mJ cm2 pulse-1
Fig. 25. (a) Number density of craters as a function of the laser power density for single-shotlaser irradiation. The
number density of
original Au NPs was (150 ± 10) particles
␮m−2 . (b) Thermoacoustic pressure estimated by the vapor pressure of gold given by p0 = (2m)
3/2
/(kB T )
1/2
03 · exp −L0 /kB T
as a function of the laser power density
of 45-nm-diameter Au NPs placed on a glass substrate (effective refractive index: 1.12). In the equation; m, 0 ; L0 ; and kB represent the atomic mass of gold; lattice vibration
frequency; sublimation energy; and Boltzmann constant, respectively. (Reproduced from ref. [116]. Reproduced with permission from the American Chemical Society.)
48
S. Hashimoto et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 13 (2012) 28–54
Fig. 26. Nanoscale-ablation site and depth profile: SEM images of the nanorods before and after laser irradiation at a local effective power density of (a) 54 mJ cm−2 , right at
the ablation threshold, and (b) 218 mJ cm−2 . The scale bars correspond to 75 nm, and the yellow arrows indicate the incident polarization. (c) The depth profile of the ablation
site shown in (b), along the long axis shown with the dotted line as obtained using AFM. Note the different scales for the horizontal and vertical axes in (c). (Reproduced from
ref. [76]. Reproduced with permission from the Optical Society of America.)
nanoholes with a diameter slightly smaller or roughly the same as
that of the spheres on a silicon surface with 250- and 40-nm Au
nanospheres exposed to a single pulse of a 532-nm nanosecond
laser [165]. These observations were attributed to the plasmonic
enhancement of the incident electric field at the boundary of the NP
and the substrate. This deduction was supported by the observation
that the nanohole profiles resembled the laser intensity distribution on the substrate surface in the power density region lower
than the ablation threshold power density of the bulk silicon. The
assumption of the plasmonic enhancement was also supported by
the fact that the holes produced by a linearly polarized wave of laser
radiation exhibited an elongated entrance shape in the direction of
polarization. Thus, besides the heat mode of enhanced material fabrication, a photon-mode process is postulated to occur, in which the
electric field enhancement due to the excitation of the LSPR band
of NPs can cause a multiphoton absorption, leading to the ablation
and modification of the material surface [160–167].
Despite demonstrations by several groups, the direct cause of
plasmon-assisted near-field ablation is still not understood very
well. Ben-Yaker’s group conducted further research to gain a better
understanding of the phenomena [76]. They observed nanoscalelaser modification and ablation of a silicon (1 0 0) surface at local
power densities significantly lower than the direct femtosecondlaser-ablation region of silicon, i.e., 420 mJ cm−2 . Fig. 26 shows SEM
images of ablation sites and a depth profile, as observed by an AFM
(atomic force microscopy).
Ben-Yaker’s group found that the nanorod ablation sites were
a photo imprint of the nanorod, very similar in size to the
nanorod. The ablation sites did not consist of two separated craters,
which is expected from a calculated |E|2 enhancement pattern. On
the basis of this observation, they concluded that the Poynting
vector enhancement more accurately predicts plasmonic-laserablation than the |E|2 enhancement pattern [76]. However, based
on the observation of a double-crater shaped hole, Meunier’s group
strongly opposed this view [168]. This dispute should be resolved
by further experimental and theoretical studies.
Ueno et al. [28,29,169–171] have demonstrated a new technique of nanogap-assisted surface plasmon nanolithography.
They prepared arrays consisting of pairs of closely spaced gold
nanoblocks on dielectric substrates including glass. They observed
the two-photon absorption (TPA)-assisted photopolymerization of
a negative-type photoresist, SU-8 in the intense near-field supposed to exist in the nanogap between the blocks generated by the
irradiation of a halogen lamp (wavelength range: 600–1000 nm,
0.2 W cm−2 ), with a CW laser (785 nm) or femtosecond laser
(800 nm) [28]. The photoresisit, SU-8 is known to photopolymerize by exposure to ultraviolet light at wavelengths shorter than
360–400 nm, and thus, the simultaneous absorption of two photons is prerequisite to initiate polymerization. Fig. 27 shows the
SEM images of pairs of Au nanoblocks, (a) with regions of polymerized SU-8 ((b) and (c)) resulting from exposure to a femtosecond
Ti:sapphaire laser (120 fs, 800 nm, 80 MHz).
When the laser beam was polarized linearly along the axis of
the pairs (Fig. 27b), longitudinal plasmon modes localized in the
nanogaps induced significant local polymerization after a short
exposure. When the laser beam was polarized perpendicular to
the axis of the pairs, transverse plasmon modes were excited
Fig. 27. (a) SEM image of a pair of gold nanoblocks measuring 100 × 100 × 40 nm3
and separated by a 5.6-nm wide nanogap before irradiation by an attenuated
femtosecond-laser beam. (b) SEM image of other nanoblock pairs after 0.01 s exposure to the laser beam polarized linearly along the long axis of the pair. (c) SEM image
of another pair after 100 s exposure to the laser beam polarized in the perpendicular
direction (d and e). Theoretically calculated near-field patterns at selected planes
for the excitation conditions of the samples shown in (b) and (c), respectively. In (d),
the field pattern is shown on the x–y plane bisecting the nanoblocks at half of their
height (i.e., 20 nm above the substrate), and in (e), the field is calculated on the plane
coincident with the line c–c shown in (d). The field intensity is normalized to that
of the incident wave, and therefore represents the intensity-enhancement factor.
(Reproduced from [28]. Reproduced with permission from the American Chemical
Society.)
S. Hashimoto et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 13 (2012) 28–54
and induced photopolymerization in the corresponding regions
(Fig. 27c). The calculated field-intensity patterns shown in Fig. 27d
and e correspond closely with the experimental photopolymerization pattern. Furthermore, for Au nanoblock structures with
nanogaps narrower than 10 nm, photopolymerization rates were
increased by orders of magnitude [169]. The same group [170,171]
also used the nanoblock structures for photomasks for contact
exposure of a positive photoresist on glass substrates. As expected,
periodic patterns of pits were formed at the nanogaps in the developed photoresist surface after exposure to a femtosecond-laser
beam. It was shown that the ordered clusters of nanoparticles separated by nanogaps can provide a stable and versatile platform for
further development of optical subwavelength nanolithography.
LSPR-assisted laser nanofabrication and nanogap-assisted surface plasmon nanolithography are promising techniques to beat
the diffraction-limited resolution (200–300 nm).
4.2. Laser-induced deposition and treatment
Niidome and co-workers [137,138,172] reported that Au NPs
can be deposited on a glass surface by using nanosecond-pulsed
532-nm laser irradiation (83 mJ cm−2 pulse−1 ) of dodecanethiolpassivated Au NPs (3.2 ± 0.95 nm) in cyclohexane. When laser
irradiation was carried out using a photomask, reddish rectangular
patterns based on the deposited Au NPs can be clearly observed by
the naked eye. The deposition of Au NPs occurs only at the laserirradiated region of the substrate. The observation is reminiscent
of the study conducted by Ito et al. [118,119] for a single Au NP (see
Section 3.2.5).
A dense monolayer of isolated Ag- and Au NPs 40–60 nm in
average diameter were prepared on mica by converting sputterdeposited Ag- and Au-island films by nanosecond-laser irradiation
at 532 nm (∼50 mJ cm−2 ) [173]. Fig. 28 shows such an example.
Besides the shape change, particle growth was observed. The resultant film gave strong LSPR bands without any serious broadening.
A relatively narrow range of laser power density was employed for
the preparation. This power density range corresponds to the particle temperatures between the melting and boiling points of Au
NPs, suggesting that particle melting triggers the transformation
observed.
In a higher laser power density regime greater or above
100 mJ cm−2 , a complicated mode of film conversion resulting in
the strong deformation (flattening) of the LSPR band became dominant before the ablation mode finally set in. The fabrication of
sphere-like Au NPs of similar shapes and alignments on sapphire,
GaN, SiO2 , and silicon substrates by irradiation of a few UV laser
pulses on a Au thin film was reported by other groups [174–176].
The nanostructural change to a periodically arranged line pattern with a modified particle size and shape distribution occurred
when thin films containing Au NPs were irradiated with femtosecond linearly polarized laser pulses (150 fs) [177]. Fig. 29 shows an
example.
A linear dependence between the period of the line structures and the laser wavelength (800, 528, 400, and 266 nm)
was observed. For the particle size and shape changes, the atomic
diffusion process (Ostwald ripening, coalescence/reshaping) was
assumed. Photothermal melting/evaporation could be the driving
force of the event. The periodic structure is reminiscent of the “ripple” or LIPSS (laser-induced periodic surface structure) that has
been formed on nearly every kind of solid material (metals, semiconductors, glasses, and polymers) [178–180]. It was assumed that
the constructive interference between incident waves and surface
waves generated by the scattering of the incident wave caused the
observed surface structural change [178]. The periodically arranged
line pattern could be an example of the self organization of Au
NPs resulting from the laser-induced photothermal effect. Here it
49
is pertinent to note that Obara and co-workers [181] presented
computational and experimental results on plasmonic and Mie
scattering control of far-field interference for regular ripple formation on semiconductor and metal. They observed the interference
ripple pattern on the silicon substrate, which originates from the
plasmonic far field by gold nanospheres irradiated by femtosecondlaser pulses. Their result suggests that in addition to plasmonic
scattering, the Mie scattering far field also play a role for the origin
of surface ripple formation.
Nanosecond- and femtosecond-laser-induced detachment was
observed for Au nanostructures prepared on a silicon or quartz
substrate by nanosphere lithography [182–184]. Laser-induced
melting was considered to be responsible for the nanosecond
excitation because of the contraction of the liquid toward a
sphere [182]. The detachment was observed for femtosecondlaser excitation in both air and liquid. The ejection mechanism
in air was considered to involve the ablation of surface atoms
from a gold particle, which generated intense pressure at the
particle–substrate interface [183]. In contrast, in a liquid environment, a mechanism was postulated that involves energy
transfer from a photoexcited nanoprism to the solvent within
cavities and defects at the particle–substrate interface [184]. In
this instance, the hot-solvent molecules result in an intense
pressure at the particle–substrate interface, resulting in particle
ejection.
4.3. Miscellaneous applications
The optical-limiting effect that strongly decreases the transmission of light at higher laser intensities was observed for Ag
nanoclusters and Au NPs dispersed in solution [185–190]. The
optical-limiting behavior is applicable to devices that protect
human eyes and solid-state sensors from intense optical beams.
Fig. 30 shows an example of a dendrimer nanocomposite of Ag
clusters, {Ag(0)}E .
The nonlinear transmission was observed at 532 nm, within the
LSPR resonance spectral range. The threshold power density for
optical limiting was 2.0 J cm−2 . The mechanism of optical limiting
was ascribed to the formation of nanobubbles. The optical-limiting
performance of {Ag(0)}E is compared well with the results obtained
with carbon nanotubes. More recent nanosecond studies expressed
rather controversial opinions [188–190]. While Polavarapu [188]
concluded that nonlinear scattering plays an important role in
the optical-limiting behavior of oleylamine-capped Au NPs, Sun’s
group [189,190] postulated the dominant contribution of freecarrier absorption for Au NP aggregates.
Because of chemical and thermal stabilities with less cytotoxicity, plasmonic NPs have been popular in biomedical applications.
In particular, the sharp molecular-scale confinement of hightemperature regions is very beneficial as a nanoheater. For
example, Takeda et al. [191,192] demonstrated that the degradation of proteins and nucleic acids can result from 532-nm
nanosecond-laser excitation (10 ns, 94 mJ cm−2 pulse−1 ) of Au NPs
an aqueous solution. For instance, they mixed Au NPs (diameter:
∼20 nm) with two kinds of protein molecules: bovine serum albumin (BSA) and lysozyme. Proteins were selectively adsorbed on Au
NPs by adjusting the pH of the solution. They observed selective
degradation because the Au NP creates a high-temperature region
in its close vicinity by the laser irradiation. At the laser power density they employed, the fragmentation/size reduction of NPs is also
likely. On the other hand, they assumed that the formation of the
laser-induced bubble is unlikely. Before these studies, a similar
study had been conducted by Huttmann et al. [193]. Under irradiation with nano- and picosecond-laser pulses, the enzymes alkaline
phosphatase and chymotrypsin bound to the surface of 15-nmdiameter Au NPs were inactivated. In this case, clear evidence of
50
S. Hashimoto et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 13 (2012) 28–54
Fig. 28. (a) AFM image (dimensions: 1000 × 1000 nm2 ) of an as-deposited Au-island film consisting of flat-shaped particles as simulated below the image. (b) 1000 × 1000 nm2
AFM image taken after laser irradiation (single-shot at ∼50 mJ cm−2 ), clearly providing evidence for conversion to good spherical nanoparticles, with a typical height profile
and size distribution shown below the image and on the right, respectively. (Reproduced from [173]. Reproduced with permission from the American Chemical Society.)
Fig. 29. (a) TEM images (upper: cross-sectional view, lower: overhead view) of a polymer film containing gold nanoparticles after laser irradiation ( = 400 nm) and (b) TEM
images of a nonirradiated polymer film containing gold nanoparticles. (Reprinted with permission from ref. [177]. Copyright 2005, American Institute of Physics.)
Fig. 30. (a) The optical-limiting behavior of the {Ag(0)}E nanocomposite at 532 nm for a 10-Hz pulse-repetition rate observed by irradiating nanosecond-laser pulses. The
solid line illustrates the case of a constant transmission of ∼60% (the transmission value at low power density). The inset shows the corresponding transmission variation
with the input power density. (b) A schematic representation of dendrimer nanocomposite of Ag clusters, {Ag(0)}E architecture. (Modified from ref. [185]. Reproduced with
permission from the American Chemical Society.)
the fragmentation was obtained by TEM measurement. However,
they reserved the judgment with regard to bubble formation as a
damage mechanism. The involvement of bubble generation with
protein deactivation is still not clear.
Another example of a biomedical approach is the application of nanobubbles to theranostics (diagnostics and therapeutics)
[194,195]. A distinct goal of theranostics is to selectively target
specific (diseased) tissues or cells to increase diagnostic and therapeutic accuracy. In an example shown in Fig. 31 [194], a single-shot
laser-pulse excitation of transient and localized vapor bubbles
(defined as PNB [45]) caused damage to living cells (lung carcinoma
cells A549) using Au NPs.
For each cell, bright field and side scattering optical images
were registered before and after the pump pulse. In Fig. 31, picture (a) shows the bright-field image before irradiation and (c)
shows it after the irradiation. Also, (b) is the scattering image
of PNBs generated inside the cell. The detection of cell damage was based on the observation of blebbing bodies in the
bright-field image of the cell after the pulse (c). The study
revealed conditions necessary for the precise excitation and
S. Hashimoto et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 13 (2012) 28–54
51
Fig. 31. In vivo generation and detection of plasmonic nanobubble (PNB) in Au NP-treated A 549 cells: (a) bright-field microscopy image of two cells prior to their exposure
to a single pump-laser pulse (532 nm, 0.5 ns, 200 mJ cm−2 ); (b) time-resolved scattering image of the intracellular PNBs obtained with the 10-ns delayed (relative to the
pump-laser pulse) pulsed-probe laser (690 nm, 0.5 ns), where the lines show the contours of the cells as in (a); (c) bright-field microscopy image of two cells obtained
120 s after their exposure to a single pump-laser pulse and the generation of PNBs; (d) time response obtained during the exposure to a single pump-laser showing bubble
expansion and collapse. (Reproduced from ref. [194]. Reproduced with permission from IOP Publishing.)
Fig. 32. Upper panel: dark-field light-scattering images of Au NP-doped zeolite L crystals. Lower panel: plots of the number of Au NPs per zeolite crystal (occupancy) versus
abundance for different quantities of Au NPs in zeolite crystals. Bar charts show experimental results for approximately 200 zeolite crystals; line plots show calculated curves
based on a Poisson distribution for various experimental average occupancies: (a) 1.8; (b) 4.2. (Reproduced from Ref. [197]. Reproduced with permission from the American
Chemical Society.)
detection of plasmonic effects in vivo around Au NPs with nanoscale
sensitivity.
Finally, we show an example of composite-material fabrication
based on laser-ablation generation of Au NPs in liquids [101–105].
Chemically prepared Au NPs are unstable in aqueous solutions containing high concentrations of salts or at extreme pHs. In zeolite
synthesis, Au NPs are required to withstand extreme high salts and
high pHs. Laser-ablation-based Au NPs were found to be very stable
in zeolite synthesis gel containing highly concentrated alkali presumably because of surface negative charge [145]. The upper panel
of Fig. 32 shows dark-field light-scattering images of Au NP-doped
zeolite L crystals [196,197]. Various colors (yellow, green, orange,
and red) of bright spots in the crystals represent Au NPs with various sizes. For instance, green, yellow, and red represent 30–50 nm,
60–80 nm, and greater than 100-nm-diameter NPs, respectively.
The measurement of single-particle scattering spectra of each spot
is feasible; the particle diameter can be estimated by the simulation of the spectra with the Mie calculation. The lower panel of
Fig. 32 represents the histograms of the number densities of Au NPs
per zeolite crystal. The number density can be altered by changing
the concentration of Au NPs in the synthesis gel. Notably, the histograms were approximated by the Poisson distribution. Beside this
example, the laser-ablation-based preparation of NPs in solution
may find many other applications.
5. Summary and future outlook
This review demonstrates the immense potential of Au NPs
when used with lasers, providing multiple varieties of research
opportunities, many of which are ongoing. The beneficial outcome
of the researches can influence various fields of science and technology: physics, chemistry, biomedicine, and materials science.
The reader may find the key word “LSPR” repeated many times
in this review. LSPR is primarily characterized not only by its distinct color but also by its highly efficient light-gathering ability
and swift/efficient light-to-heat conversion capability. Regarding
the mechanistic side, both ultrafast femtosecond and conventional
nanosecond lasers contributed enormously to reveal the dynamics
52
S. Hashimoto et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 13 (2012) 28–54
and processes initiated by the excitation of Au NPs. Although these
dynamics have been largely unraveled for the ensemble of particles because of the effort of many researchers, investigations to
develop better understanding of the single-particle level with better time resolution is still needed. Additionally, bubble dynamics is
currently the most intriguing subject and needs significant effort
to reveal the complete picture. From the perspective of materials,
we observed the powerful nature of pulsed lasers in combination with Au NPs. Both constructive and destructive effects were
demonstrated. The former can be the photothermal fabrication of
homogeneous spherical particles, NP monolayers on substrates,
periodic structures, and composite materials. Here heat management is crucial to successful fabrication. The latter might be the
explosive evaporation, spallation, and fragmentation of Au NPs
together with bubble formation. However, in addition to already
established destruction of malignant cells, this can be exploited
for the fabrication of nanoholes. Photothermal bubbles have
potential use in material fabrication and diagnostic/therapeutic
applications. Because of remarkably high-resolution properties,
plasmon-assisted nanofabrication and nanolithography continue
to be a subject of intense research. We will undoubtedly observe
further activities and many beneficial results in the laser-plasmonic
NP research field for many years to come.
Acknowledgements
KAKENHI (no. 23310065 and no. 22655043) to S. H. from the
Japan Society for the Promotion of Science (JSPS) is gratefully
acknowledged. JSPS fellowship to D. W. is gratefully acknowledged.
Professor Hiroshi Masuhara is acknowledged for his continuous
support and encouragement.
References
[1] U. Kreibig, M. Vollmer, Optical Properties of Metal Clusters, Springer, Berlin,
1995.
[2] P. Mulvaney, Langmuir 12 (1996) 788–800.
[3] C.F. Bohren, D.R. Huffman, Absorption and Scattering of Light by Small Particles, Wiley, New York, 1983.
[4] S. Link, M.A. El-Sayed, J. Phys. Chem. B 103 (1999) 8410–8426.
[5] S. Link, M.A. El-Sayed, Int. Rev. Phys. Chem. 19 (2000) 409–453.
[6] G.V. Hartland, Chem. Rev. 111 (2011) 3858–3887.
[7] C. Burda, X. Chen, R. Narayanan, M.A. El-Sayed, Chem. Rev. 105 (2005)
1025–1102.
[8] S.A. Maier, M.L. Brongersma, P.G. Kik, S. Meltzer, A.A.G. Requicha, Adv. Mater.
13 (2001) 1501–1505.
[9] W.L. Barnes, A. Dereux, T.W. Ebbesen, Nature 424 (2003) 824–830.
[10] K. Sokolov, M. Follen, J. Aaron, I. Pavlova, A. Malpica, R. Lotan, R. RichardsKortum, Cancer Res. 63 (2003) 1999–2004.
[11] C.J. Murphy, A.M. Gole, J.W. Stone, P.N. Sisco, A.M. Alkilany, E.C. Goldsmith,
S.C. Baxter, Acc. Chem. Res. 41 (2008) 1721–1730.
[12] S. Lal, S.E. Clare, N.J. Halas, Acc. Chem. Res. 41 (2008) 1842.
[13] X. Huang, P.K. Jain, I.H. El-Sayed, M.A. El-Sayed, Laser Med. Sci. 23 (2008)
217–228.
[14] J. Kneipp, H. Kneipp, K. Kneipp, Chem. Soc. Rev. 37 (2008) 1052–1160.
[15] L. Brus, Acc. Chem. Res. 41 (2008) 1742–1749.
[16] K.A. Willets, R.P. Van Duyne, Annu. Rev. Phys. Chem. 58 (2007) 267–297.
[17] J.N. Anker, W.P. Hall, O. Lyandres, N.C. Shah, J. Zhao, R.P. Van Duyne, Nat.
Mater. 7 (2008) 442–453.
[18] K.M. Mayer, J.H. Hafner, Chem. Rev. 111 (2011) 3828–3857.
[19] M. Graetzel, J. Photochem. Photobiol. C Rev. 4 (2003) 145–153.
[20] A.C. Mayer, S.R. Scully, B.E. Hardin, M.W. Rowell, M.D. McCehee, Mater. Today
10 (2007) 28–33.
[21] P.V. Kamat, J. Phys. Chem. C 112 (2008) 18737–18753.
[22] H.A. Atwater, A. Polman, Nat. Mater. 9 (2010) 205–213.
[23] S. Gao, K. Uno, H. Misawa, Acc. Chem. Res. 44 (2011) 251–260.
[24] P. Bharadwaj, B. Deutsch, L. Novotny, Adv. Opt. Photon. 1 (2009) 438–483.
[25] V. Giannini, A.I. Fernandez-Dominguez, S.C. Heck, S.A. Maier, Chem. Rev. 111
(2011) 3888–3921.
[26] I. Cohanoshi, S. Yao, K.D. Belfield, F.E. Hernandez, J. Appl. Phys. 101 (2007)
086112.
[27] Y. Tsuboi, R. Shimizu, T. Shoji, N. Kitamura, J. Am. Chem. Soc. 131 (2009)
12623–12627.
[28] K. Ueno, S. Juodkazis, T. Shibuya, Y. Yokota, V. Mizekis, K. Sasaki, H. Misawa,
J. Am. Chem. Soc. 130 (2008) 6928–6929.
[29] K. Ueno, S. Juodkazis, V. Mizekis, K. Sasaki, H. Misawa, Adv. Mater. 20 (2008)
26–30.
[30] K. Aslan, I. Gryczynski, J. Malicka, E. Matveeva, J.R. Lakowicz, C.D. Geddes, Curr.
Opin. Biotech. 16 (2006) 55–62.
[31] J.R. Lakowicz, Plasmonics 1 (2006) 5–33.
[32] Y. Tian, T. Tatsuma, J. Am. Chem. Soc. 127 (2010) 2344–7637.
[33] E. Kowalska, B. Ohtani, Phys. Chem. Chem. Phys. 12 (2010) 4965–4971.
[34] J. Hohlfeld, S.-S. Wellershoff, J. Gudde, U. Conrad, V. Jahnke, E. Matthias, Chem.
Phys. 251 (2000) 237–258.
[35] C. Bauer, J.-P. Abid, D. Fermin, H.H. Girault, J. Chem. Phys. 120 (2004)
9302–9315.
[36] A.L. Tchebotareva, P.V. Ruijgrok, P. Zijlstra, M. Orrit, Laser Photonics Rev. 4
(2010) 581–597.
[37] G.V. Hartland, Annu. Rev. Phys. Chem. 57 (2006) 403–430.
[38] S. Link, M.B. Burda, B. Mohamed, B. Nikoobakht, M.A. El-Sayed, J. Phys. Chem.
A 103 (1999) 1165–1170.
[39] S. Link, M.B. Burda, B. Nikoobakht, M.A. El-Sayed, J. Phys. Chem. B 104 (2000)
6152–6163.
[40] R.R. Letfullin, C. Joenathan, T.F. George, V.P. Zharov, Nanomedicine 1 (2006)
473–480.
[41] V.K. Pustovalov, A.S. Smetannikov, V.P. Zharov, Laser Phys. Lett. 5 (2008)
775–792.
[42] V. Kotaidis, A. Plech, Appl. Phys. Lett. 87 (2005) 213102.
[43] V. Kotaidis, C. Dahmen, G. von Plessen, F. Springer, A. Plech, J. Chem. Phys. 124
(2006) 184702.
[44] J. Neumann, R. Brinkmann, J. Biomed. Opt. 10 (2005) 024001.
[45] D. Lapotko, Opt. Express 17 (2009) 2538–2556.
[46] J.M. Sun, B.S. Gerstman, B. Li, J. Appl. Phys. 88 (2000) 2352–2362.
[47] X. Wang, X. Xu, J. Therm. Stresses 25 (2002) 457–473.
[48] X. Michalet, F.F. Pinaud, L.A. Bentolila, J.M. Tsay, S. Doose, J.J. Li, G. Sundaresan,
A.M. Wu, S.S. Gambhir, S. Weiss, Science 307 (2005) 538–544.
[49] V. Biju, T. Itoh, M. Ishikawa, Chem. Soc. Rev. 39 (2010) 3031–3056.
[50] J.P. Wilcoxon, J.E. Martin, F. Parsapour, B. Wiedenman, D.F. Kelley, J. Chem.
Phys. 108 (1998) 9137–9143.
Yguerabide,
E.E.
Yguerabide,
Anal.
Biochem.
262
(1998)
[51] J.
137–156.
[52] S. Schultz, D.R. Smith, J.J. Mock, D.A. Schultz, Proc. Natl. Acad. Sci. 97 (2000)
996–1001.
[53] M. Rycenga, C.M. Cobley, J. Zeng, W. Li, C.H. Moran, Q. Zhang, D. Qin, Y. Xia,
Chem. Rev. 111 (2011) 3669.
[54] M.-C. Daniel, D. Astruc, Chem. Rev. 104 (2004) 293–346.
[55] H. Nabika, M. Takase, F. Nagasawa, K. Murakoshi, J. Phys. Chem. Lett. 1 (2010)
2470–2487.
[56] P.B. Johnson, R.W. Christy, Phys. Rev. B 6 (1972) 4370–4379.
Mie,
Ann.
Phys.
25
(1908)
377–445.
[57] G.
German
version:
http://diogenes.iwt.uni-bremen.de/vt/laser/papers/Mie-Beitrag-zurOptik-AnnalenPhysik1908.PDF (accessed January 1, 2012), English version:
http://diogenes.iwt.uni-bremen.de/vt/laser/papers/SAND78-6018-Mie1908-translation.pdf (accessed January 1, 2012).
[58] A computational program, MiePlot: http://www.philiplaven.com/mieplot.htm
(accessed January 1, 2012) can be conveniently used for the calculation of
Cext, Cabs, and Csca.
[59] C. Soennichsen, T. Franzl, T. Wilk, G. Von Pressen, J. Feldmann, New J. Phys. 4
(2002) 93.
[60] A. Arbouet, D. Christofilos, N. Del Fatti, F. Vallee, J.R. Huntzinger, L. Arnaaud,
P. Billaud, M. Boyer, Phys. Rev. Lett. 93 (2004) 127401.
[61] K.S. Lee, M.A. El-Sayed, J. Phys. Chem. B 109 (2005) 20331–20338.
[62] P.K. Jain, K.S. Lee, I.H. El-Sayed, M.A. El-Sayed, J. Phys. Chem. B 110 (2006)
7238–7248.
[63] B.J. Wiley, S.H. Im, Z.-Y. Li, J. McLellan, A. Siekkinen, Y. Xia, J. Phys. Chem. B
110 (2006) 15666–15675.
[64] M. Otter, Z. Phys. 161 (1961) 539–549.
[65] S. Link, M.A. El-Sayed, J. Phys. Chem. B 103 (1999) 4212–4217.
[66] T.S. Ahmadi, S.L. Logunov, M.A. El-Sayed, J. Phys. Chem. 100 (1996)
8053–8056.
[67] S.L. Logunov, T.S. Ahmadi, M.A. El-Sayed, J.T. Khoury, R.L. Whetten, J. Phys.
Chem. B 101 (1997) 3713–3719.
[68] S. Link, C. Burda, Z.L. Wang, M.A. El-Sayed, J. Chem. Phys. 111 (1999)
1255–1264.
[69] O. Ekici, R.K. Harrison, N.J. Durr, D.S. Eversole, M. Lee, A. Ben-Yakar, J. Phys. D.
Appl. Phys. 41 (2008) 185501.
[70] D. Werner, A. Furube, T. Okamoto, S. Hashimoto, J. Phys. Chem. C 115 (2011)
8503–8512.
[71] L. Anisimov, B.L. Kapeliovich, T.L. Perel’sman, Soviet Phys. JETP 39 (1974)
375–377.
[72] C.K. Sun, F. Vallee, L.H. Acioli, E.P. Ippen, J.G. Fujimoto, Phys. Rev. B 50 (1994)
15337–15348.
[73] F. Caupin, E. Herbert, C. R. Phys. 7 (2006) 1000–1017.
[74] B.J. Messinger, K. Ulrich, R.K. Chang, P.W. Barber, Phys. Rev. B 24 (1981)
649–657.
[75] C.F. Bohren, Am. J. Phys. 51 (1983) 323–326.
[76] R.K. Harrison, A. Ben-Yaker, Opt. Express 18 (2010) 22556–22571.
[77] J. Boneberg, J. Koenig-Birk, H.-J. Munzer, P. Leiderer, K.L. Shuford, G.C. Schatz,
Appl. Phys. A 89 (2007) 299–303.
[78] K.L. Kelly, E. Coronado, L.L. Zhao, G.C. Schatz, J. Phys. Chem. B 107 (2003)
668–677.
S. Hashimoto et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 13 (2012) 28–54
[79] P.R. Sajanlal, C. Subramaniam, P. Sasanpour, B. Rashidian, T. Pradeep, J. Mater.
Chem. 20 (2010) 2108–2113.
[80] S-S. Chang, C-W. Shih, C-D. Chen, W-C. Lai, C.R. Chris Wang, Langmuir 15
(1999) 701–709.
[81] S. Link, M.A. El-Sayed, J. Chem. Phys. 114 (2001) 2362–2368.
[82] P. Zijlstra, J.W.M. Chon, M. Gu, Phys. Chem. Chem. Phys. 11 (2009) 5915–5921.
[83] Ph. Buffat, J-P. Borel, Phys. Rev. A 13 (1976) 2287–2298.
[84] K. Dick, T. Dhanasekaran, Z. Zhang, D. Meisel, J. Am. Chem. Soc. 124 (2002)
2312–2317.
[85] K.K. Nanda, Pramana J. Phys. 72 (2009) 617–628.
[86] S. Inasawa, M. Sugiyama, Y. Yamaguchi, J. Phys. Chem. B 109 (2005)
3104–3111.
[87] H. Petrova, J.P. Juste, I. Pastoriza-Santos, G.V. Hartland, L.M. Liz-Marzan, P.
Mulvaney, Phys. Chem. Chem. Phys. 8 (2006) 814–821.
[88] A. Plech, R. Cerna, V. Kotaidis, F. Hudert, A. Bartels, T. Dekorsy, Nano Lett. 7
(2007) 1026–1031.
[89] A. Plech, V. Kotaidis, C.S. Gresillon, C. Dahmen, G. von Plessen, Phys. Rev. B 70
(2004) 195423.
[90] A. Takami, H. Kurita, S. Koda, J. Phys. Chem. B 103 (1999) 1226–1232.
[91] S. Inasawa, M. Sugiyama, Y. Yamaguchi, J. Phys. Chem. B 109 (2005)
9404–9410.
[92] A. Pyatenko, M. Yamaguchi, M. Suzuki, J. Phys. Chem. C 113 (2009) 9078–9085.
[93] D. Werner, S. Hashimoto, T. Uwada, Langmuir 26 (2010) 9956–9963.
[94] P.V. Kamat, M. Fluminani, G.V. Hartland, J. Phys. Chem. B 102 (1998) 3123.
[95] K. Yamada, Y. Tokumoto, T. Nagata, F. Mafune, J. Phys. Chem. C 110 (2006)
11751–11756.
[96] K. Yamada, K. Miyajima, F. Mafune, J. Phys. Chem. C 111 (2007) 11246–11251.
[97] M. Shoji, K. Miyajima, F. Mafune, J. Phys. Chem. C 112 (2008) 1929–1932.
[98] H. Muto, K. Miyajima, F. Mafune, J. Phys. Chem. C 112 (2008) 5810–5815.
[99] F. Giammanco, E. Giorgetti, P. Marsili, A. Giusti, J. Phys. Chem. C 114 (2010)
3354–3363.
[100] D. Werner, S. Hashimoto, J. Phys. Chem. C 115 (2011) 5063–5072.
[101] G.W. Wang, Prog. Mater. Sci. 52 (2007) 648–698.
[102] S. Barcikowski, F. Devasa, K. Moldenhauer, J. Nanopart. Res. 11 (2009)
1883–1893.
[103] V. Amendola, M. Meneghetti, Phys. Chem. Chem. Phys. 11 (2009) 3805–3821.
[104] S. Sakamoto, M. Fujutsuka, T. Majima, J. Photochem. Photobiol. C Rev. 10
(2009) 33–56.
[105] N.G. Semaltianos, Crit. Rev. Solid State Mater. Sci. 35 (2010) 105–124.
[106] A. Plech, V. Kotaidis, M. Lorenc, J. Bonberg, Nat. Phys. 2 (2006) 44–47.
[107] J.H. Hodak, I. Martini, G.V. Hartland, J. Phys. Chem. B 102 (1998) 6958–6967.
[108] J.H. Hodak, A. Henglein, G.V. Hartland, J. Chem. Phys. 111 (1999) 8613–
8621.
[109] P.B. Corkum, F. Brunel, N.K. Sherman, Phys. Rev. Lett. 61 (1988) 2886–2889.
[110] B.N. Chichkov, C. Momma, S. Nolte, F. Von Alvensleben, A.A. Tuennermann,
Appl. Phys. A Mater. Sci. Process. 63 (1996) 109–115.
[111] W.A. Saunders, Phys. Rev. A 46 (1992) 7028–7041.
[112] U. Naeher, S. Bjornholm, S. Frauendorf, F. Garcias, C. Guet, Phys. Rep. 285
(1997) 245–320.
[113] Z. Lin, L.V. Zhigilei, Phys. Rev. B 77 (2008) 075133.
[114] L. Rayleigh, Philos. Mag. 5 (1882) 184–187.
[115] C. Novo, P. Mulvaney, Nano. Lett. 7 (2007) 520–524.
[116] S. Hashimoto, T. Uwada, M. Hagiri, R. Shiraishi, J. Phys. Chem. C 115 (2011)
4986–4993.
[117] O. Warshavski, L. Minai, G. Bisker, D. Yelin, J. Phys. Chem. C 115 (2011)
3910–3917.
[118] S. Ito, H. Yoshikawa, H. Masuhara, Appl. Phys. Lett. 80 (2002) 482–484.
[119] S. Ito, T. Mizuno, H. Yoshikawa, H. Masuhara, Jpn. J. Appl. Phys. 46 (2007)
L241–L243.
[120] A. Ashkin, Phys. Rev. Lett. 24 (1970) 156–159.
[121] K. Svoboda, S.M. Block, Opt. Lett. 19 (1994) 930–932.
[122] Y. Tsuboi, T. Shoji, N. Kitamura, M. Takase, K. Murakoshi, Y. Mizumoto, H.
Ishihara, J. Phys. Chem. Lett. 1 (2010) 2327–2333.
[123] M. Hu, G.V. Hartland, J. Phys. Chem. B 106 (2002) 7029–7033.
[124] S. Merabia, S. Shenogin, L. Joly, P. Keblinski, J-L. Barrat, Proc. Natl. Acad. Sci.
106 (2009) 15113–15118.
[125] V. Juve, M. Scardamaglia, P. Malioli, A. Crut, S. Marabia, L. Joly, N. Del Fatti,
Phys. Rev. B 80 (2009) 195406.
[126] A. Plech, V. Kotaidis, M. Lorenc, M. Wulff, Chem. Phys. Lett. 401 (2005)
565–569.
[127] A. Siems, S.S.L. Weber, J. Boneberg, A. Plech, New J. Phys. 13 (2011) 043018.
[128] E.Y. Lukianova-Hleb, D.O. Lapotko, Nano Lett. 9 (2009) 2160–2166.
[129] E. Lukianova-Hleb, Y. Hu, L. Latterini, L. Tarpani, S. Lee, R.A. Drezek, J.H. Hafner,
D.O. Lapotko, ACS Nano 4 (2010) 2109–2123.
[130] E.Y. Lukianova-Hleb, L.J.E. Anderson, S. Lee, J.H. Hafner, D.O. Lapotko, Phys.
Chem. Chem. Phys. 12 (2010) 12237–12244.
[131] J. Neumann, R. Brinkmann, J. Appl. Phys. 101 (2007) 114701.
[132] J. Neunmann, R. Brinkmann, Appl. Phys. Lett. 93 (2008) 033901.
[133] H. Fujiwara, S. Yanagida, P.V. Kamat, J. Phys. Chem. B 103 (1999) 2589–2591.
[134] C.S. Ah, H.S. Han, K. Kim, D-J. Jang, Pure Appl. Chem. 72 (2000) 91–99.
[135] Z. Peng, T. Walther, K. Kleinermanns, Langmuir 21 (2005) 4249–4253.
[136] Y. Niidome, A. Hori, T. Sata, S. Yamada, Chem. Lett. 29 (2000) 310–311.
[137] H. Takahashi, Y. Niidome, T. Sato, S. Yamada, Colloid Surf. A 247 (2004)
105–113.
[138] T. Kawasawa, H. Takahashi, H. Kawasaki, Y. Niidome, S. Yamada, Trans. Mater.
Res. Soc. Jpn. 30 (2005) 571–574.
53
[139] D. Werner, S. Hashimoto, T. Tomita, S. Matsuo, Y. Makita, J. Phys. Chem. C 112
(2008) 16801–16808.
[140] H. Wang, A. Piatenko, K. Kawaguchi, X. Li, Z. Swiatkowska-Warkocka, N.
Koshizaki, Angew. Chem. Int. Ed. 49 (2010) 6361–6364.
[141] X. Li, A. Piatenko, Y. Shimizu, H. Wang, K. Koga, N. Koshizaki, Langmuir 27
(2011) 5076–5080.
[142] H. Wang, N. Koshizaki, L. Li, L. Jia, K. Kawaguchi, X. Li, A. Piatenko,
Z. Swiatkowska-Warkocka, Y. Bando, D. Golberg, Adv. Mater. 23 (2011)
1865–1870.
[143] X. Li, N. Koshizaki, A. Piatenko, Y. Shimizu, H. Wang, J. Liu, X. Wang, M. Gao,
Z. Wang, X. Zeng, Opt. Express 19 (2011) 12855.
[144] S. Pocovi-Martinez, M. Parreno-Romero, S. Agouram, J. Perez-Prieto, Langmuir
27 (2011) 5234–5241.
[145] J-P. Sylvestre, S. Poulin, A.V. Kabashin, E. Sacher, M. Meunier, J.H.T. Loung, J.
Phys. Chem. B 108 (2004) 16864–16869.
[146] H. Hasegawa, N. Saitoh, K. Tsuji, K. Kimura, Z. Phys. D. 20 (1991) 325–327.
[147] N. Saitoh, H. Hasegawa, K. Tsujii, K. Kimura, J. Phys. Chem. 98 (1994)
2143–2147.
[148] K. Kimura, J. Phys. Chem. 98 (1994) 11997–12002.
[149] K. Kimura, Bull. Chem. Soc. Jpn. 69 (1996) 321–324.
[150] Y. Takeuchi, T. Ida, K. Kimura, J. Phys. Chem. B 101 (1997) 1322–1327.
[151] H. Eckstein, U. Kreibig, Z. Phys. D. 26 (1993) 239–241.
[152] S.V. Karpov, V.V. Slabko, G.A. Chiganova, Colloid J. 64 (2002) 425–441.
[153] H. Yoshikawa, T. Matsui, H. Masuhara, Phys. Rev. E 70 (2004) 061406.
[154] Y. Tanaka, H. Yoshikawa, T. Itoh, M. Ishikawa, Opt. Express 17 (2009)
18760–18767.
[155] Y. Zhang, C. Gu, A.M. Schwartzberg, S. Chen, J.Z. Zhang, Phys. Rev. B 73 (2006)
165405.
[156] S.H. Ko, Y. Choi, D.J. Hwang, C.P. Grigoropoulos, Appl. Phys. Lett. 89 (2006)
141126.
[157] K. Yamada, T. Itoh, Y. Tsuboi, Appl. Phys. Express 1 (2008) 087001.
[158] S. Hashimoto, T. Uwada, M. Hagiri, H. Takai, T. Ueki, J. Phys. Chem. C 113 (2009)
20640–20647.
[159] K. Furukawa, T. Kameyama, K. Okazaki, T. Yako, M. Otsuki, Y. Tsuboi, T. Torimoto, Chem. Lett. 40 (2011) 1411–1413.
[160] N.N. Nedyalkov, H. Takada, M. Obara, Appl. Phys. A 85 (2006) 163–168.
[161] N.N. Nedyalkov, T. Sakai, T. Mianishi, M. Obara, J. Phys. D. Appl. Phys. 39 (2006)
5037–5042.
[162] N.N. Nedyalkov, P.A. Atanasov, M. Obara, Nanotechnology 18 (2007) 305703.
[163] P. Leiderer, C. Bartels, J. Konig-Birk, M. Mosbacher, J. Boneberg, Appl. Phys.
Lett. 85 (2004) 5370–5372.
[164] A. Kolloch, T. Geldhauser, K. Ueno, H. Misawa, J. Boneberg, A. Plech, P. Leider,
Appl. Phys. A. 104 (2011) 793–799.
[165] A. Heltzel, S. Theppakuttai, S.C. Chen, J.R. Howell, Nanotechnology 19 (2008)
025305.
[166] D. Eversole, B. Luk’yanchuk, A. Ben-Yaker, Appl. Phys. A 89 (2007) 283–291.
[167] A. Plech, P. Leiderer, J. Boneberg, Laser Photon Rev. 3 (2009) 435–451.
[168] E. Boulais, A. Robitalle, P. Desjeans-Gauthier, M. Meunier, Opt. Express 19
(2011) 6177–6178.
[169] K. Ueno, S. Juodkazis, T. Shibuya, V. Mizeikis, Y. Yokota, H. Misawa, J. Phys.
Chem. C 113 (2009) 11720–117249.
[170] K. Ueno, S. Takabatake, Y. Nishijima, V. Mizeikis, Y. Yokota, H. Misawa, J. Phys.
Chem. Lett. 1 (2010) 657–662.
[171] K. Ueno, S. Takabatake, Y. Nishijima, V. Mizeikis, Y. Yokota, H. Misawa, Appl.
Phys. Lett. 99 (2011) 011107.
[172] Y. Niidome, A. Hori, H. Yakahashi, Y. Goto, S. Yamada, Nano Lett. 1 (2001)
365–369.
[173] M. Kawasaki, M. Hori, J. Phys. Chem. B 107 (2003) 6760–6765.
[174] C-Y. Chen, J-Y. Fu-Ji Tsai, Y-C. Lu, Y-W. Kiang, C.C. Yang, Opt. Express 17 (2009)
14186–14198.
[175] C-H. Lin, L. Zhou, H. Xiao, S-J. Chen, H-L. Tsai, Opt. Lett. 35 (2010) 941–943.
[176] D.-Q. Yang, M. Meunier, E. Sacher, J. Appl. Phys. 95 (2004) 5023–5026.
[177] A. Keisow, S. Strojkark, K. Loeschner, A. Heilmann, A. Poldipensky, A.
Abdolvand, G. Seifert, Appl. Phys. Lett. 86 (2005) 153111.
[178] J.E. Sipe, J.F. Young, J.S. Preston, H.M. van Driel, Phys. Rev. B 27 (1983)
1141–1154.
[179] Y. Shimotsuma, P.G. Kazansky, J. Qui, K. Hirao, Phys. Rev. Lett. 91 (2003)
247405.
[180] M. Hung, F. Zhao, Y. Cheng, N. Xu, Z. Xu, ACS Nano 3 (2009) 4062–4070.
[181] G. Obara, N. Maeda, T. Miyanishi, M. Terakawa, N.N. Nedyalkov, M. Obara, Opt.
Express 19 (2011) 19093–19103.
[182] A. Habenicht, M. Olapinski, F. Burmeister, P. Leiderer, J. Boneberg, Science 309
(2005) 2043–2045.
[183] W. Huang, W. Qian, M.A. El-Sayed, J. Am. Chem. Soc. 128 (2006) 13330–13331.
[184] C. Tabor, W. Qian, M.A. El-Sayed, J. Phys. Chem. C 111 (2007) 8934–8941.
[185] R.G. Ispasoiu, L. Balough, O.P. Varnavski, D.A. Tomalia, T. Goodson III, J. Am.
Chem. Soc. 122 (2000) 11005–11006.
[186] L. François, M. Mostafavi, J. Belloni, J-F. Delouis, J. Delaire, P. Feneyrou, J. Phys.
Chem. B 104 (2000) 6133–6137.
[187] L. François, M. Mostafavi, J. Belloni, J.A. Delaire, Phys. Chem. Chem. Phys. 3
(2001) 4965–4971.
[188] L. Polavarapu, N. Venkatram, W. Ji, Q.-H. Xu, ACS Appl. Mater. Interface 1
(2009) 2298–2303.
[189] W. Sun, Q. Dai, J.G. Worden, Q. Huo, J. Phys. Chem. B 109 (2005) 20854–20857.
[190] G. Wang, W. Sun, J. Phys. Chem. B 110 (2006) 20901–20905.
[191] Y. Takeda, F. Mafune, T. Kondow, J. Phys. Chem. C 113 (2009) 5027–5030.
54
S. Hashimoto et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 13 (2012) 28–54
[192] Y. Takeda, T. Kondow, F. Mafune, Phys. Chem. Chem. Phys. 13 (2011) 586–592.
[193] G. Huttmann, B. Radt, J. Serbin, R. Birngruber, Proc. SPIE 5142 (2003)
88–95.
[194] E.Y. Lukianova-Helb, C. Santiago, D.S. Wagner, J.H. Hafner, D.O. Lapotko, Nanotechnology 21 (2010) 225102.
[195] E.Y. Lukianova-Helb, E.Y. Hanna, J.H. Hafner, D.O. Lapotko, Nanotechnology
21 (2010) 085102.
[196] S. Hashimoto, T. Uwada, H. Masuhara, T. Asahi, J. Phys. Chem. C 112 (2008)
15089–15093.
[197] S. Hashimoto, J. Phys. Chem. Lett. 2 (2011) 509–519.